The core challenge in multi‑agent coordination under hostile environments is to derive policy inference mechanisms that remain reliable when agents’ observations are subtly perturbed by adversaries. Adversarial observation perturbations (AOPs) can stem from noisy telemetry, malicious sensor spoofing, or targeted semantic manipulation (e.g., prompt injection in LLM‑driven agents). The objective is therefore to construct inference frameworks that can (i) detect, (ii) adapt to, and (iii) recover fr...
The objective of this chapter is to articulate a trust‑aware federated aggregation framework that can be deployed across heterogeneous multi‑agent networks—such as fleets of UAVs, edge IoT nodes, autonomous vehicles, and industrial cyber‑physical systems—while simultaneously guaranteeing:1. Integrity and robustness of the global model against data‑poisoning, Byzantine, and targeted adversarial updates.2. Privacy preservation through differential privacy and secure, verifiable aggregation.3...
The primary objective of this chapter is to articulate a forward‑looking blueprint for resilient interpretability in adversarial multi‑agent systems, specifically targeting the threat of communication sabotage. In environments where agents must coordinate under partial observability, malicious actors can inject deceptive messages, corrupt shared beliefs, or silently hijack coordination protocols. We seek to develop a principled, theory‑of‑mind (ToM)‑driven defense architecture that (1) detects a...
The central challenge addressed in this chapter is the allocation of a finite explainability budget—the computational, human, and regulatory resources dedicated to interpreting model decisions—so as to maximize sample‑efficiency in resilient, adversarial multi‑agent reinforcement learning (MARL) systems. In high‑stakes domains such as autonomous logistics, finance, and healthcare, agents must learn from limited interactions while remaining interpretable to satisfy regulatory mandates and stakeho...
The objective of this chapter is to articulate a forward‑looking framework that amplifies misalignment signals arising from partial observability in multi‑agent reinforcement learning (MARL) systems, thereby enabling resilient interpretability and trustworthy coordination. Specifically, we aim to:1. Quantify how incomplete state information inflates credit‑assignment and coordination errors;2. Develop abstraction‑driven representations that preserve task‑relevant modalities while filtering s...
The goal is to design a gradient‑masking strategy that simultaneously enhances adversarial robustness and maintains, or even improves, the interpretability of deep multi‑agent AI systems. In a coordinated setting, agents must not only withstand adversarial perturbations but also provide transparent, trustworthy explanations of their decisions to human operators and regulatory bodies. Traditional masking methods often obscure gradients enough to mislead attackers but at the cost of rendering sali...
The central research challenge is to develop counterfactual explanation (CE) mechanisms that remain faithful, actionable, and interpretable when subjected to adversarial perturbations—both input‑level noise and model‑level shifts. Existing CE methods exhibit brittleness: perturbations that flip a model’s prediction are often treated as noisy artifacts rather than actionable changes, leading to misleading explanations and compromised user trust. Our objective is to bridge the gap between the opti...
The objective of this chapter is to articulate a systematic approach for resilient blame attribution within cooperative multi‑agent systems (MAS) that are deployed in adversarial or partially‑observable environments. Specifically, we aim to:1. Identify how misattribution of blame undermines coordination, trust, and safety in MAS;2. Survey the prevailing conventions for blame assignment and their limitations;3. Propose a frontier framework that couples causal attribution, counterfactual rea...
In multi‑agent AI systems that coordinate under uncertainty, a pervasive problem is the cascading misinterpretation of local signals that propagates through the network, leading to suboptimal joint actions. The objective of this chapter is to synthesize the state of the art on how interpretability gaps, noisy communications, and adversarial perturbations jointly degrade coordination, and to propose a frontier methodology that explicitly couples joint interpretability with adaptive trust to break...
The central goal of this chapter is to prevent explainability models from over‑fitting to benign data while operating within adversarial multi‑agent AI systems. In coordinated agent settings, explanations must remain faithful when the environment is perturbed—whether by intentional adversarial attacks, distribution shift, or evolving agent policies. Over‑fitting leads to brittle explanations that fail to surface hidden biases or to reveal the true decision logic under malicious conditions, there...
The goal of this chapter is to articulate a forward‑looking blueprint that transforms the way multi‑agent AI systems retrieve, validate, and interpret information in the presence of adversarial threats. Specifically, we seek to:1. Mitigate knowledge‑base corruption (e.g., poisoned documents, membership inference leaks, and unauthorized content injection).2. Guarantee interpretability and traceability of each retrieved fact, enabling agents to audit and explain their reasoning.3. Enable res...
The central challenge addressed in this chapter is the amplification of hallucinated content within collaborative multi‑agent deliberations. As autonomous agents increasingly coordinate through structured debate, the very mechanisms designed to surface truth—repeated argumentation, cross‑checking, and voting—can paradoxically propagate false claims when agents echo each other or succumb to sycophancy. The objective is to delineate the conditions under which hallucination amplification occurs, re...
The chapter seeks to delineate a research agenda that transitions from conventional defensive practices against prompt‑level attacks to a frontier framework capable of detecting, interpreting, and neutralizing deceptive explanations generated by large‑language and multimodal systems. In particular, we aim to:1. Characterize how adversarial prompt injections can induce misleading chain‑of‑thought (CoT) narratives that conceal illicit intent.2. Integrate mechanistic interpretability and indepe...
The primary objective of this chapter is to delineate the susceptibility of multi‑agent system (MAS) communication graphs to malicious actors and to chart a research trajectory that transitions from traditional resilience techniques to frontier‑grade, adaptive defense architectures. We seek to:1. Quantify how graph‑structural properties (degree, robustness, connectivity) influence the spread of adversarial influence.2. Expose the failure modes of existing consensus protocols (e.g., W‑MSR) wh...
The central challenge is to construct a resilient, interpretable multi‑agent AI (MAIA) framework that can maintain reliable coordination under hostile, dynamic, and uncertain environments. In operational domains such as autonomous UAV swarms, cyber‑physical sensor networks, and decentralized financial systems, adversaries may inject false data, poison training streams, or subvert inter‑agent communication protocols to disrupt mission objectives or compromise safety. The objective is therefore tw...
| [v9] | Concept-Guided Fine-Tuning: Steering ViTs away from Spurious Correlations to Improve Robustness https://arxiv.org/abs/2603.08309 Model performance is typically contrasted with in-distribution accuracy on standard benchmarks like ImageNet and its variants (ImageNet-v2 ).Our work evaluates extensively on these OOD datasets to demonstrate meaningful improvements in robustness. ... |
| [v46] | Decentralized Multi-Agent Swarms for Autonomous Grid Security in Industrial IoT: A Consensus-based Approach https://doi.org/10.48550/arXiv.2601.17303 CVT combines Byzantine fault-tolerant consensus protocols with domain-specific threat scoring via a weighted voting system that accounts for each agent's accuracy and the proximity of its threat to its own threat assessment. CVT achieves sub-millise... |
| [v81] | Federated microservices architecture with blockchain for privacy-preserving and scalable healthcare analytics https://doi.org/10.1038/s41598-026-39837-1 Blockchain's immutable ledger and smart contract capabilities have been explored for healthcare auditability and data integrity. Kumar et al. surveyed blockchain-integrated federated learning in edge-fog-cloud healthcare applications, highlighting se... |
| [v84] | Pipeline monitoring data recovery using novel deep learning models: an engineering case study https://pubmed.ncbi.nlm.nih.gov/41127626/ The model integrates three components: the prairie dog optimization algorithm (PDO) for hyperparameter tuning, the bidirectional gated recurrent unit (BiGRU) for effective temporal feature extraction, and the generative adversarial network (GAN) for ... |
| [v92] | State-of-the-Art Deep Learning Methods for Microscopic Image Segmentation: Applications to Cells, Nuclei, and Tissues https://doi.org/10.3390/jimaging10120311 The system demonstrates significant performance improvements, with cross-magnification MAP increasing from 0.313 to 0.551, and a 15.68% boost in cross-domain adaptability. Overall, FARS effectively delivers reliable predictions in medical image analy... |
| [v114] | A Bayesian Framework for Uncertainty-Aware Explanations in Power Quality Disturbance Classification https://arxiv.org/abs/2604.13658 Second, each posterior sample θ (s) simultaneously generates a predictive sample f θ (s) (x) and an explanation sample R (s) (x), thereby coupling predictive and explanation uncertainty through shared posterior draws.This structural parallel with Bay... |
| [v299] | D3HRL: A Distributed Hierarchical Reinforcement Learning Approach Based on Causal Discovery and Spurious Correlation Detection https://doi.org/10.48550/arxiv.2505.01979 Sample-efficient goal-conditioned reinforcement learning via predictive information bottleneck for goal representation learning. Q Zou, E Suzuki, 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE2023 Highly valued subgoal ge... |
| [v385] | AI brings clear opportunity and real risk. https://www.softwareimprovementgroup.com/blog/iso-standards-for-ai/ ISO and IEC publish a coherent set of standards that cover AI concepts, lifecycle engineering, risk management, governance and quality. Start with the items below to structure your program and your audits. Purpose in your AI program ISO/IEC 42001:2... |
| [v448] | 2019 AI Alignment Literature Review and Charity Comparison (Larks) (summarized by Rohin): As in three previous years (AN #38), this mammoth post goes through the work done within AI alignment from De https://www.lesswrong.com/s/dT7CKGXwq9vt76CeX/p/D7CY29s2D6HJirqcF Adversarial imitation learning seeks to avoid this by training a discriminator reward model with the agent: the discriminator is trained via supervised learning to distinguish between expert trajectories and agent trajectories, while the agent tries ... |
| [v461] | ONG: One-Shot NMF-based Gradient Masking for Efficient Model Sparsification https://arxiv.org/abs/2508.12891 Deep Neural Networks (DNNs) have achieved remarkable success but their large size poses deployment challenges. While various pruning techniques exist, many involve complex iterative processes, specialized criteria, or struggle to maintain sparsity ef... |
| [v478] | The transition from simple Large Language Model (LLM) calls to autonomous AI agents represents a paradigm shift in software engineering. https://dev.to/kuldeep_paul/top-10-metrics-to-monitor-for-reliable-ai-agent-performance-4b36 In Retrieval Augmented Generation (RAG) systems, this is often measured as ""Faithfulness"": is the answer derived strictly from the retrieved context? Why it matters: In domains like healthcare, finance, or legal, a hallucination is a liability. H... |
| [v511] | Reducing inference cost of Alzheimer's disease identification using an uncertainty-aware ensemble of uni-modal and multi-modal learners https://pubmed.ncbi.nlm.nih.gov/39952976/ We propose a novel MRI- and FDG PET-based multi-modal deep learning approach that mimics clinical decision-making by incorporating uncertainty estimates of an MRI-based model (generated using Monte Carlo dropout and evidential deep learning) to deter... |
| [v547] | RAL2M: Retrieval Augmented Learning-To-Match Against Hallucination in Compliance-Guaranteed Service Systems https://doi.org/10.48550/arXiv.2601.02917 To our knowledge, this work is the first to systematically study LLMs for query matching with a focus on hallucination mitigation, formulating the Retrieval-Augmented Learningto-Match problem for LLM deployment with zero-generation hallucination in c... |
| [v570] | Facilitates the identification of counterfactual queries in structural causal models via the ID* and IDC* algorithms by Shpitser, I. and Pearl, J. (2007, 2008) , . http://cran.ma.ic.ac.uk/web/packages/cfid/index.html Construction of parallel worlds graphs and counterfactual graphs is carried out automatically based on the counterfactual query and the causal diagram. See Tikka, S. (2023) for a tutorial of the package. Suggests: covr, dagitty, igraph, mockery, tes... |
| [v577] | Neurosymbolic Framework for Concept-Driven Logical Reasoning in Skeleton-Based Human Action Recognition https://arxiv.org/abs/2605.07140 Our framework bridges representation learning and symbolic inference by grounding first-order logic predicates in learnable spatial and temporal motion concepts. Specifically, we employ a standard spatio-temporal skeleton encoder to extract latent mo... |
| [v625] | Stability-Driven Motion Generation for Object-Guided Human-Human Co-Manipulation https://arxiv.org/abs/2604.20336 Our results (d) maintain coordinated grasps and stable payload alignment, whereas previous methods exhibit slipping contacts or delayed responses when the green object changes its pose. Figure 5 .Figure 6 . 56 Figure 5. Cooperative motions produce... |
| [v647] | Secure Pipelines, Smarter AI: LLM-Powered Data Engineering for Threat Detection and Compliance https://www.preprints.org/manuscript/202504.1365 When combined, they can support audit trails, selective data masking, and fine-grained control policies that satisfy both technical and legal scrutiny . The hybrid compliance layer enhances not only governance but also explainability. While LLMs enr... |
| [v654] | Efficient Domain Coverage for Vehicles with Second-Order Dynamics via Multi-Agent Reinforcement Learning https://doi.org/10.48550/arxiv.2211.05952 However, designing model-based controllers is challenging, and the state-of-the-art classical control policy still exhibits a large degree of sub-optimality. In this paper, we present a reinforcement learning (RL) approach for the multi-agent efficie... |
| [v675] | InterAgent: Physics-based Multi-agent Command Execution via Diffusion on Interaction Graphs https://doi.org/10.48550/arXiv.2512.07410 We further propose a novel interaction graph exteroception representation that explicitly captures fine-grained joint-to-joint spatial dependencies to facilitate network learning. Additionally, within it we devise a sparse edge-based attention mechan... |
| [v676] | Multi-agent Communication with Graph Information Bottleneck under Limited Bandwidth https://www.semanticscholar.org/paper/de7e81b1c897c85e0bc88e6644ece43bcac06c4f Based on the above discussion, in this paper, we focus on the problem of bandwidth-constrained communication in MARL. To simultaneously address the challenges of whom to communicate with and what to communicate, we propose a novel and universal multi... |
| [v696] | State-Action Inpainting Diffuser for Continuous Control with Delay https://arxiv.org/abs/2603.01553 The fundamental limitation of explicit belief estimation lies in the nature of the regression task involved in continuous control.Unlike classification, where decision boundaries can be robust to minor perturbations, continuous state regression is hi... |
| [v722] | Learning-Based Resource Allocation Scheme for TDD-Based CRAN System https://arxiv.org/abs/1608.07949 However, for time division duplex (TDD) MIMO systems, the resource allocation is done based on instantaneous CSI availability (without using learning, or considering the CSI acquistion overhead), where resource allocation is referred to RB assignment... |
| [v758] | Maintainer: Hans W. Borchers <[email protected]> https://cran.asia/web/packages/pracma/refman/pracma.html B.A. Pearlmutter, Fast Exact Multiplication by the Hessian, Neural Computation (1994), Vol. 6, Issue 1, pp.... |
| [v804] | A Loss Curvature Perspective on Training Instability in Deep Learning https://arxiv.org/abs/2110.04369 Lanczos algorithm only requires Hessian-vector products which can be efficiently computed via Pearlmutter's trick . (2021)... |
| [v821] | The rapid advancements in AI, particularly the release of large language models (LLMs) and their applications, have attracted significant global interest and raised substantial concerns on responsibl http://www.wikicfp.com/cfp/servlet/event.showcfp These AI systems, especially autonomous LLM agents and those involving multi-agent interacting, require careful system-level engineering to ensure responsible AI and AI safety. In recent years, numerous regulations, principles, and guidelines for re... |
| [v867] | Essentially no human intervention': Chinese AI solves 12-year-old math problem in just 80 hours - and even proves it https://www.techradar.com/pro/essentially-no-human-intervention-chinese-ai-solves-12-year-old-math-problem-in-just-80-hours-and-even-proves-it Similarly, proofs produced by large language models are prone to hallucination and are far less reliable than formal verification methods. The Chinese team's framework bridges the gap between natural language reasoning and formal machine verificatio... |
| [v869] | IT Security News Daily Summary 2026-03-13 https://www.itsecuritynews.info/it-security-news-daily-summary-2026-03-13/ Linux Servers to Full Root Takeover 7:2 : Authorities Disrupt SocksEscort Proxy Botnet Exploiting 369,000 IPs Across 163 Countries 6:36 : New Critical MediaTek Vulnerability Exposes Android Phone PINs to Theft in 45 seconds 6:36 : RSAC Innovation ... |
| [v885] | authID Unveils Mandate Framework to Establish the Critical Trust and Governance Layer for the Accelerating Agentic AI Market https://www.businesswire.com/news/home/20251118838387/en/authID-Unveils-Mandate-Framework-to-Establish-the-Critical-Trust-and-Governance-Layer-for-the-Accelerating-Agentic-AI-Market Mandate defines how organizations establish accountability for autonomous activity: each agent is sponsored by a verified human so that it operates within explicitly authorized boundaries, and the platform produces immutable records that can be audit... |
| [v888] | Cyber-Resilient Perception: Safeguarding Autonomous Vehicles With Trust-Aware Sensor Fusion https://doi.org/10.1109/sr.2025.3562156 This study developed a trust-aware sensor fusion framework to enhance AV resilience against cyber-physical attacks.By leveraging Dirichlet trust distributions, real-time anomaly detection, and cross-sensor consistency checks, the system dynamically r... |
| [v903] | Robotic fleet management systems are increasingly vital for sustainable operations in agriculture, forestry, and other field domains where labor shortages, efficiency, and environmental concerns inte https://www.frontiersin.org/journals/robotics-and-ai/articles/10.3389/frobt.2025.1706910/full A central design principle of FORMIGA is the standardisation of communication between heterogeneous agents - robots and humans - through the Robot Operating System (ROS). ROS provides a flexible framework for modular robot software, and in FORMIGA it... |
| [v909] | Understanding Generalization through Decision Pattern Shift https://arxiv.org/abs/2605.13148 Empirical analyses across multiple datasets and architectures show that, (i) decision patterns form a highly structured, class-consistent space with strong intra-class cohesion and low inter-class confusion, enabling direct analysis of a model's deci... |
| [v923] | Pass Your Professional Google Workspace Administrator Exams - 100% Money Back Guarantee! https://www.test-king.com/cert-Professional-Google-Workspace-Administrator.htm Administrators are often required to connect Google Workspace with other identity providers, cloud services, or third-party applications. Candidates should gain familiarity with SAML, OAuth, and API access configurations. Practical exercises may incl... |
| [v947] | LLM as Graph Kernel: Rethinking Message Passing on Text-Rich Graphs https://arxiv.org/abs/2603.14937 GAT (Velickovic et al., 2017).A type of GNN with attention weights to differentiate neighbor importance during aggregation.This design improves robustness to noisy neighbors, making GAT a representative example of graph models that enhance aggregatio... |
| [v959] | The Role of Blockchain in Zero Trust Architecture | HackerNoon https://hackernoon.com/the-role-of-blockchain-in-zero-trust-architecture Blockchain complements Zero Trust in several critical ways. First, it can store user and device credentials in a manner that makes tampering exceedingly difficult. Where traditional identity systems rely on centralized databases, a blockchain-based i... |
| [v962] | XXX-X-XXXX-XXXX-X/XX/$XX. https://doi.org/10.48550/arxiv.2306.06071 We evaluate the impact of various adversarial attacks on the accuracy of YOLOv5, including L-BFGS, FGSM, C&W, BIM, PGD, One Pixel Attack, and Universal Adversarial Perturbations attack . This paper aims to identify and analyze the effect of such atta... |
| [v995] | Frequency-Aware Model Parameter Explorer: A new attribution method for improving explainability https://doi.org/10.48550/arXiv.2510.03245 Gradient-based techniques, including Saliency Maps (SM), Grad-CAM , and Score-CAM , improved interpretability but lacked fine-grained Figure 1: An illustration of frequency filtering. The top row displays an image separated into its low-frequency (bl... |
| [v1010] | ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks https://aclanthology.org/2024.findings-naacl.85/ %T ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks... |
| [v1026] | This edition consolidates and stabilizes the generative integration first formalized in PSRT v2.0, and supersedes the earlier PTI-focused v1. https://zenodo.org/records/17932629 Process → Structure → Recursion (PSTR) In PSRT v2.1, this generative identity is formally acknowledged but operationally constrained. The framework adopts the bounded formulation: PSRT v2.1 = UTI PTI HPE subject to the Unified Failure Domain (UFD)... |
| [v1039] | Prior to Liverpool, I worked at the University of Oxford, the University of New South Wales, and the Chinese Academy of Sciences. https://cgi.csc.liv.ac.uk/~xiaowei/ We also consider verification of both robustness and resilience [Neurocomputing, 2024], as well as extending robustness verification to the deep reinforcement learning [RA-L, 2024]. We extend randomised smoothing technique to reinforcement learning ... |
| [v1040] | CAFED-Net: Cross-Adaptive Federated Learning with Dynamic Adversarial Defence for Real-Time Privacy-Preserving and Threat Detection in Distributed IoT Ecosystems https://doi.org/10.30880/jscdm.2025.06.01.004 Their detecting power and the ability to adapt to the simulation-based assessment, however, prove to be more effective than the baseline mode ls in the circumstances that occur under adversarial drift. In this study, the authors introduce a solution ... |
| [v1043] | Hierarchical Task Network Planning for Facilitating Cooperative Multi-Agent Reinforcement Learning https://doi.org/10.48550/arxiv.2306.08359 Current MARL approaches often fail to to learn policies effectively in this multi-agent setting due to the joint actions of agents affecting the multi-agent system and the lack of non-zero reward drive. To address this issue, one way is to abstract ... |
| [v1048] | Recently, deep multi-agent reinforcement learning (MARL) has gained significant popularity due to its success in various cooperative multi-agent tasks. https://doi.org/10.48550/arxiv.2308.11272 A fully cooperative multi-agent task can be seen as a decentralized partially observable Markov decision process (Dec-POMDP) (Oliehoek and Amato 2016), represented as a tuple G = ⟨S, A, P, r, Z, O, O, I, n, γ⟩.... |
| [v1052] | Total Accepted Paper Count 2670 http://deepnlp.org/content/paper/nips2022 Most existing approaches find such attributions either using activations and gradients or by repeatedly perturbing the input. We instead address this challenge by training a second deep network, the Explainer, to predict attributions for a pre-traine... |
| [v1080] | Bipedal Action Model For Humanoid Robot https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260126805).pn The co-trained of the combined L2/L1 model can be an end-to-end process, where the error between the L1 model's predicted action and a ground-truth demonstration are backpropagated through both models. This allows the high-level L2 model to be fine-t... |
| [v1172] | Hybrid Reputation Aggregation: A Robust Defense Mechanism for Adversarial Federated Learning in 5G and Edge Network Environments https://doi.org/10.1109/OJCOMS.2025.3646134 Our ablation studies further demonstrate that the full hybrid system achieves 98.66% accuracy, while the anomaly-only and reputation-only variants drop to 84.77% and 78.52%, respectively, validating the synergistic value of our dual-mechanism approac... |
| [v1211] | Beyond Semantic Relevance: Counterfactual Risk Minimization for Robust Retrieval-Augmented Generation https://arxiv.org/abs/2605.01302 Grounded in causal intervention, we introduce a Cognitive Perturbation Protocol to simulate user biases during training, which is then distilled into a lightweight Evidence Critic. This scoring module learns to identify documents that possess suffici... |
| [v1220] | Submitted on 18 Feb 2025 (v1), last revised 3 Sep 2025 (this version, v2)] https://arxiv.org/abs/2502.12616 To achieve a trade-off, this paper investigates methods to disentangle content from logical reasoning without a complete formalisation. In particular, we present QuaSAR (for Quasi-Symbolic Abstract Reasoning), a variation of CoT that guides LLMs to o... |
| [v1259] | When you're coordinating multiple ai agents on one task, how do you keep them from breaking the handoffs? - https://community.latenode.com/t/when-youre-coordinating-multiple-ai-agents-on-one-task-how-do-you-keep-them-from-breaking-the-handoffs/60678 If it doesn't match, validation fails and you have a clear error, not a silent misinterpretation. The coordination works when you eliminate ambiguity upfront, not when you rely on the AI to figure it out. PixelPioneer88 January 23, 2026, 9:38pm 5 ... |
| [v1321] | The "Awakening Moment" for Agents: EverOS Brand Upgrade and Public Beta Launches the Era of Self-Evolving Memory - Laotian Times https://laotiantimes.com/2026/04/14/the-awakening-moment-for-agents-everos-brand-upgrade-and-public-beta-launches-the-era-of-self-evolving-memory/ It natively parses and stores diverse data types (PDFs, images, Word docs, spreadsheets, URLs) via a single API. Its hybrid retrieval fuses dense semantic vectors, sparse keyword matching, and multimodal alignment, ensuring that agents can accurately... |
| [v1334] | Online Bayesian system identification in multivariate autoregressive models via message passing https://arxiv.org/abs/2506.02710 N Ta, M Kobilarov, F Dellaert, International Conference on Unmanned Aircraft Systems. IEEE2014 Linear optimal control on factor graphs-a message passing perspective. C Hoffmann, P Rostalski, IFAC-PapersOnLine. 5012017 A unifying view of estimation ... |
| [v1346] | HawkEye 360, Inc.: 424B4 (424B4) https://www.sec.gov/Archives/edgar/data/0001628280/0001628280-26-032207-index.htm Our customers face ongoing adversarial threats in active conflicts and require real-time situational awareness across the signal spectrum. Customers increasingly demand rapid, actionable data, edge autonomy, and cost-effective mission solutions. Trad... |
| [v1355] | FlowSteer: Guiding Few-Step Image Synthesis with Authentic Trajectories https://arxiv.org/abs/2511.18834 Our Online Trajectory Alignment (OTA) resolves both problems by training on authentic teacher trajectories, ensuring the teacher operates on-trajectory and training matches inference distributions.Adversarial distillation on trajectory Adversarial di... |
| [v1365] | One moment, a coin's soaring like a rocket, the next it's plumbing the depths, all within hours. https://digitalfinancenews.com/technology/mastering-crypto-pair-trading-with-rl/ A model trained exclusively on bull market data will likely struggle, or even fail, during a bear market. It's like training a racehorse only on flat tracks and then expecting it to win a steeplechase! This necessitates continuous monitoring and oft... |
| [v1372] | Build production RAG that actually works at scale. https://blog.premai.io/building-production-rag-architecture-chunking-evaluation-monitoring-2026-guide/ Pure vector (dense) retrieval misses exact-match queries. BM25 (sparse) retrieval misses semantic queries.... |
| [v1592] | A Resilient Distributed Algorithm for Solving Linear Equations https://doi.org/10.1109/cdc49753.2023.10383841 Resilient constrained consensus has been partially solved in only for complete graphs and studied in with an incomplete proof.It is worth emphasizing that discrete-time constrained consensus, first proposed in , in general does not enjoy exponentia... |
| [v1679] | Medical Reasoning in the Era of LLMs: A Systematic Review of Enhancement Techniques and Applications https://doi.org/10.48550/arXiv.2508.00669 Closing the "accountabil-ity gap" (Habli et al., 2020) requires a robust framework built on shared responsibility policies for developers and institutions (Information Technology Industry Council, 2024), inherently auditable and explainable AI system... |
| [v1806] | Yet its opaque "black boxes" raise serious concerns in high - stakes domains like credit, trading, fraud detection, and risk compliance. https://www.infosecured.ai/i/banking-security/explainable-ai-in-finance/ Preferred tools: LIME and SHAP dominate alongside feature - importance and rule - based methods, with hybrid multi - method frameworks growing in popularity. Deficits and challenges: lack of standard evaluation metrics, insufficient user - targeted ... |
| [v1835] | Structure and position-aware graph neural network for airway labeling - NewsBreak https://www.newsbreak.com/news/2484286429231/structure-and-position-aware-graph-neural-network-for-airway-labeling Finally, a substantial set of experiments is reported to evaluate the performance of the algorithms and support the theoretical findings. The obtained results show that the proposed strategies approximate the theoretical distance for samples close to... |
| [v1880] | Adversarial Hallucination Engineering: Targeted Misdirection Attacks Against LLM Powered Security Operations Centers https://doi.org/10.20944/preprints202512.0913.v1 Large Language Models (LLMs) are increasingly deployed in Security Operations Centers (SOCs) for alert triage and threat - intelligence synthesis. We study Adversarial Hallucination Engineering (AHE): attacks that bias LLM reasoning by introducing sm... |
| [v1909] | RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards https://doi.org/10.48550/arXiv.2506.07736 Its structure includes (1) 333,963 question-answer samples annotated with risk meta-labels spanning 14 harm types, and (2) 361,903 preference-based comparisons independently rating responses on helpfulness and harmlessness. Derived from over 16,000 a... |
| [v1915] | In 2025, public rules meet production reality: the EU AI Act sets penalties up to 7% of global turnover for certain violations, while customers expect transparent systems that show their work. https://themortonreport.com/blog/trustworthy-ai-a-step-by-step-guide-to-reliable-transparent-systems/ Maintain an AI bill of materials that lists model versions, datasets, third-party components, and licenses. For suppliers, request security attestations and evaluation summaries, and plan tests to validate claims before integration. ISO/IEC 42001:20... |
| [v1977] | Counterfactual Explanations with Probabilistic Guarantees on their Robustness to Model Change https://arxiv.org/abs/2408.04842 Abstract: Counterfactual explanations (CFEs) guide users on how to adjust inputs to machine learning models to achieve desired outputs. While existing research primarily addresses static scenarios, real-world applications often involve data or model ... |
| [v2010] | Democratizing ML for Enterprise Security: A Self-Sustained Attack Detection Framework https://doi.org/10.48550/arxiv.2512.08802 Furthermore, LLMpowered agents show promise in improving the explainability of detection results and adapting to novel, zero-day attacks, which traditionally suffer from a lack of historical data . In dynamic threat environments, security models req... |
| [v2014] | Overfitting occurs when an AI model becomes so tightly tuned to its training dataset that it begins to "memorize" its noise, quirks, and outliers rather than learning generalizable patterns. https://www.c-sharpcorner.com/article/overfitting-in-ai-why-data-governance-is-the-key-to-smarter-more-reliable-mode/ This oversight is crucial for avoiding the trap of "high accuracy" masking deeper flaws, such as overfitting, bias, or unethical decision-making. 4) Prevention Strategies Through Combined Governance Common technical strategies to reduce overfitting... |
| [v2016] | DRIFT: Divergent Response in Filtered Transformations for Robust Adversarial Defense https://arxiv.org/abs/2509.24359 To assess whether our defense induces masking artifacts, we visualize the loss surface around input x along two random, orthonormal directions (u, v) in input space, plotting L(x + au + bv, y), (a, b) ∈ 2 , on a 41 41 grid with τ = 3/255.For stochas... |
| [v2044] | Agentic AI Framework for Smart Inventory Replenishment https://doi.org/10.48550/arXiv.2511.23366 Jannelli et al. presented the agentic collaboration described by the authors as using LLM, which entails making consensusbased procurement decisions with the help of natural language arguments, which is a breakthrough in the direction of autonomous ... |
| [v2060] | The Architectural Evolution of Intelligence: A Formal Taxonomy of the AI Technology Stack https://www.c-sharpcorner.com/article/the-architectural-evolution-of-intelligence-a-formal-taxonomy-of-the-ai-technol/ The World Wide Web Consortium (W3C) standards stack comprising the Resource Description Framework (RDF), RDF Schema (RDFS), and the Web Ontology Language (OWL) provides a mathematically grounded apparatus for representing entities, their properties, ... |
| [v2111] | What Is Agentic AI in Regulatory Operations? https://www.freyrsolutions.com/what-is-agentic-ai-in-regulatory-operations Improved Audit Readiness: Maintains detailed audit trails and documentation aligned with regional and global authorities. Operational Efficiency: Reduces manual workload in regulatory affairs teams by up to 65%, freeing experts to focus on strategic... |
| [v2138] | Clinical Implementation of Artificial Intelligence in Endoscopy: A Human-Artificial Intelligence Interaction Perspective https://pubmed.ncbi.nlm.nih.gov/41572653/ Regardless of the AI capabilities, the visualization quality and systematic inspection remain fundamental prerequisites, and traditional apprenticeship training cannot be replaced by technology. This review examines AI implementation in endoscopy fro... |
| [v2147] | DUE: Dynamic Uncertainty-Aware Explanation Supervision via 3D Imputation https://doi.org/10.1145/3637528.3671641 Oring et al. proposed a regularization method that molds the latent space into a smooth, locally convex manifold consistent with training images. presents a method for interpolating between generative models of the StyleGAN architecture in a resolut... |
| [v2168] | Provenance Verification of AI-Generated Images via a Perceptual Hash Registry Anchored on Blockchain https://doi.org/10.48550/arXiv.2602.02412 Future work could explore infrastructure-level interoperability, including shared governance models, standardized registry interfaces, or common cryptographic primitives, while maintaining strict separation between content provenance and identity ver... |
| [v2173] | Byzantine Robust Cooperative Multi-Agent Reinforcement Learning as a Bayesian Game https://doi.org/10.48550/arXiv.2305.12872 In this study, we explore the robustness of cooperative multi-agent reinforcement learning (c-MARL) against Byzantine failures, where any agent can enact arbitrary, worst-case actions due to malfunction or adversarial attack. To address the uncertain... |
| [v2261] | Enhancing Network Intrusion Detection Systems: A Real-time Adaptive Machine Learning Approach for Adversarial Packet-Mutation Mitigation https://doi.org/10.1109/NCA61908.2024.00042 We introduce an Adaptive Layered Mutation Algorithm (ALMA) for generating advanced adversarial examples and a runtime adaptive learning framework for real-time detection and response.... |
| [v2277] | This is just a glorified webhook wrapper around existing API calls. https://news.ysimulator.run/item/7241 If one AI in the chain misreads intent or optimizes for the wrong objective, the user may not know until after the workspace has been altered. The real risk is not malicious use but emergent behavior in a system where responsibility is distributed an... |
| [v2296] | HEXAR: a Hierarchical Explainability Architecture for Robots https://arxiv.org/abs/2601.03070 Finally, after executing f e , ∀e ∈ E s , the explainer selector must aggregate the set of explanations {x e |e ∈ E s } into a single explanation x if |E s | > 1. The aggregation method may be implemented in a number of ways, for example, using an ... |
| [v2306] | Large Language Models (LLMs) are revolutionary, but they have a fundamental limitation: their knowledge is frozen in time. https://www.remio.ai/post/rag-vs-cag-the-ultimate-guide-to-choosing-your-ai-s-knowledge-strategy-in-2026 As the model processes this information, it creates an internal state representation from each of its self-attention layers. This captured state is called the Key-Value Cache, or KV Cache. The KV Cache is the model's encoded, digested form of your en... |
| [v2309] | F5 is a channel-led business, and we want to be crystal clear: the acquisition of CalypsoAI benefits our partners as much as it does our customers. https://www.f5.com/fr_fr/company/blog/q-and-a-with-lisa-citron-what-does-the-calypsoai-acquisition-mean-for-f5-partners Using adversarial attack simulation backed by the preeminent AI threat library, generating over 10,000 attack prompts per month, partners can deliver detailed insights for identifying vulnerabilities in real time. Furthermore, partners can help cust... |
| [v2406] | One strategy: Deploy GPT-5.2 for reasoning (100% AIME), Claude for coding (80.9% SWE-bench), Gemini Flash for speed (3x faster), Llama 4 for privacy (self-hosted), DeepSeek for scale (27x cheaper). https://www.adwaitx.com/ai-implementation-guide-2026-models-tools/ One strategy: Deploy GPT-5.2 for reasoning (100% AIME), Claude for coding (80.9% SWE-bench), Gemini Flash for speed (3x faster), Llama 4 for privacy (self-hosted), DeepSeek for scale (27x cheaper). ... The breakthrough feature of 2026 models is adjus... |
| [v2439] | Less is More: Robust Zero-Communication 3D Pursuit-Evasion via Representational Parsimony https://arxiv.org/abs/2603.08273 Abstract: Asymmetric 3D pursuit-evasion in cluttered voxel environments is difficult under communication latency, partial observability, and nonholonomic maneuver limits. While many MARL methods rely on richer inter-agent coupling or centralized sign... |
| [v2514] | Sentra-Guard: A Real-Time Multilingual Defense Against Adversarial LLM Prompts https://arxiv.org/abs/2510.22628 Abstract: This paper presents a real-time modular defense system named Sentra-Guard. The system detects and mitigates jailbreak and prompt injection attacks targeting large language models (LLMs). The framework uses a hybrid architecture with FAISS-i... |
| [v2529] | InFoBERT: Zero-Shot Approach to Natural Language Understanding Using Contextualized Word Embedding https://doi.org/10.26615/978-954-452-072-4_025 Jian-Guo Zhang, Kazuma Hashimoto, Chien-Sheng Wu, Yao Wan, S Philip, Richard Yu, Caiming Socher, Xiong, arXiv:1910.03544arXiv preprintJian-Guo Zhang, Kazuma Hashimoto, Chien-Sheng Wu, Yao Wan, Philip S Yu, Richard Socher, and Caiming Xiong. 2019. Fin... |
| [v2577] | Trustworthy Orchestration Artificial Intelligence by the Ten Criteria with Control-Plane Governance https://doi.org/10.48550/arXiv.2512.10304 However, the standard operates at the management level without prescribing architectural properties that AI systems must exhibit, particularly for orchestrated, multi-component ecosystems where governance must be enforced as a runtime property rather... |
| [v2615] | OgbujiPT is a general-purpose knowledge bank system for LLM-based applications. https://pypi.org/project/OgbujiPT/ It provides a unified API for storing, retrieving, and managing semantic knowledge across multiple backends, with support for dense vector search, sparse retrieval, hybrid search, and more.... |
| [v2616] | Regulation of algorithms https://en.wikipedia.org/?curid=63442371 The GDPR's policy on the right of citizens to receive an explanation for algorithmic decisions highlights the pressing importance of human interpretability in algorithm design. In 2016, China published a position paper questioning the adequacy of exi... |
| [v2655] | Constrained Optimal Fuel Consumption of HEVs under Observational Noise https://arxiv.org/abs/2410.20913 Z Lin, G Thomas, G Yang, T Ma, Advances in Neural Information Processing Systems. 202033173 Maximum entropy rl (provably) solves some robust rl problems. B Eysenbach, S Levine, arXiv:2103.062572021arXiv preprint Robust reinforcement learning as a s... |
| [v2689] | In an era where autonomous machines and connected systems are becoming integral to daily life, the question of how these systems can trust one another moves from theoretical curiosity to practical i https://bioengineer.org/building-trust-a-new-framework-to-enhance-safety-in-robot-and-vehicle-networks/ Beyond laboratory studies, the research underscores the urgent need to embed cy-trust principles into policy and regulatory frameworks, particularly as autonomous systems rapidly transition from controlled environments to public domains. Cities are a... |
| [v2810] | Agents Under Siege: Breaking Pragmatic Multi-Agent LLM Systems with Optimized Prompt Attacks https://doi.org/10.18653/v1/2025.acl-long.476 Our goal is to systematically vary the underlying communication structure, so we can quantify the impact of network topology on adversarial robustness.Experimental details are listed in Appendix B.4 The results for the ablation are summarized in Fig... |
| [v2828] | Originally when Clado was first started when it was still called Linkd, there was one database for each school with approximately 10k profiles per school. https://www.davidbshan.com/writings/building-sota-people-search Agentic chunking experiments: using LLMs to summarize each profile into multiple semantic facets. Hybrid retrieval (sparse + dense): evaluating Milvus BM25 + vector hybrid search, and why query-term explosion and large-scale union merges became proh... |
| [v2830] | Controllable Stylistic Text Generation with Train-Time Attribute-Regularized Diffusion https://arxiv.org/abs/2510.06386 Improving diffusion models inverse problems using manifold constraints. Hyungjin Chung, Byeongsu Sim, Dohoon Ryu, Jong Chul, Ye , Advances in Neural Information Processing Systems. 202235 Diffusion models beat gans on image synthesis. Prafulla Dhari... |
| [v2853] | Posted on Mar 23 Originally published at blckalpaca. https://dev.to/blckalpaca/llm-landscape-2026-the-enterprise-decision-guide-eu-compliant-153l The DACH region faces particularly complex challenges: EU AI Act high-risk obligations take effect August 2026, GDPR enforcement for AI is intensifying, and German, Austrian, and Swiss regulators are each building distinct national frameworks. The 2... |
| [v2861] | Modeling eye gaze velocity trajectories using GANs with spectral loss for enhanced fidelity https://doi.org/10.1038/s41598-025-05286-5 This study introduces a Generative Adversarial Network (GAN) framework employing Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN) generators and discriminators to generate high-fidelity synthetic eye gaze velocity trajectories. We... |
| [v2879] | MAGIC-MASK: Multi-Agent Guided Inter-Agent Collaboration with Mask-Based Explainability for Reinforcement Learning https://arxiv.org/abs/2510.00274 Agents use it to steer exploration by deprioritizing perturbations in states that are visually or semantically similar to those already marked as critical by peers, which reduces redundancy and increases behavioural diversity. The protocol operates i... |
| [v2884] | The era of asking a single chatbot a question and receiving a static response is rapidly coming to an end. https://fueler.io/blog/the-complete-guide-to-multi-agent-systems-in-artificial-intelligence Increased Execution Time and Latency: Because multi-agent workflows involve multiple steps and decision-making gates, they take longer to complete than single queries, which can be a drawback for applications requiring instant responses. Why it matt... |
| [v2937] | Second Order Optimization for Adversarial Robustness and Interpretability https://arxiv.org/abs/2009.04923 The condition that the Hessian of the loss, H, be positive semi-definite has been shown to hold locally for all x, excluding a set of measure 0, when the network uses ReLU activations and the loss is categorical cross entropy (Singla et al. 2019). C... |
| [v2941] | Performance-Aware Self-Configurable Multi-Agent Networks: A Distributed Submodular Approach for Simultaneous Coordination and Network Design https://doi.org/10.48550/arxiv.2409.01411 But ActionCoordination incurs a suboptimality cost C({N i } i∈N ) due to requiring the agents to coordinate exchanging local information only, prohibiting also multi-hop communication, in favor of decision speed.For this reason, given the agents' ban... |
| [v2988] | Federated Learning Paper in Conferences https://github.com/weimingwill/awesome-federated-learning/blob/master/conferences.md Towards Model Agnostic Federated Learning Using Knowledge Distillation Diurnal or Nocturnal? Federated Learning of Multi-branch Networks from Periodically Shifting Distributions Recycling Model Updates in Federated Learning: Are Gradient Subspaces Lo... |
| [v3006] | Multi-model assurance analysis showing large language models are highly vulnerable to adversarial hallucination attacks during clinical decision support https://pubmed.ncbi.nlm.nih.gov/40753316/ We embedded fabricated content in clinical prompts to elicit adversarial hallucination attacks in multiple large language models.... |
| [v3192] | Time Series Forecasting with Missing Data Using Generative Adversarial Networks and Bayesian Inference https://doi.org/10.3390/info15040222 We propose a novel framework that combines the strengths of Generative Adversarial Networks (GANs) and Bayesian inference.... |
| [v3219] | Which prompting technique can protect against prompt injection attacks? https://www.ace4sure.com/aif-c01/which-prompting-technique-can-protect-against-prompt-question-answer.html Adversarial prompting helps uncover and mitigate these risks before deployment. Explanation of other options: B. Zero-shot prompting provides no examples and does not protect against injection attacks. C. Least-to-most prompting is a reasoning tec... |
| [v3255] | Multi-Agent Reinforcement Learning (MARL) is a rapidly evolving field that promises dynamic solutions for complex tasks within multi-agent systems (MAS) 1. https://atoms.dev/insights/multi-agent-reinforcement-learning-for-coding-foundations-applications-challenges-and-future-directions/2d27a831498a42fb91e22937bd6b95fc Interpretability and Explainability: Ensuring that the actions and recommendations of MARL agents are understandable and transparent to human developers is crucial for trust and effective collaboration . Further work is needed to trace decisions in c... |
| [v3261] | Pruning the parameters of deep neural networks has generated intense interest due to potential savings in time, memory and energy both during training and at test time. https://aiqianji.com/blog/article/4013 GraSP is a more recent algorithm that aims to preserve gradient flow at initialization by scoring weights based on the Hessian-gradient product.... |
| [v3333] | Generalized Per-Agent Advantage Estimation for Multi-Agent Policy Optimization https://arxiv.org/abs/2603.02654 This scheme improves credit assignment for off-policy trajectories by balancing sensitivity to the agent's own policy changes with robustness to non-stationarity from other agents. Experiments on benchmarks demonstrate that our approach outperforms e... |
| [v3338] | Abstract: AI safety and alignment research has predominantly been focused on methods for safeguarding individual AI systems, resting on the assumption of an eventual emergence of a monolithic AGI. Th https://www.emergentmind.com/papers/2512.16856 Reputation system manipulation: No formal model of collusion-resilient and gaming-resistant reputation; develop aggregation rules, decay functions, and anomaly detectors robust to strategic rating attacks and venue-hopping. Collusion detection (expl... |
| [v3355] | Multi-Stakeholder Alignment in LLM-Powered Collaborative AI Systems: A Multi-Agent Framework for Intelligent Tutoring https://doi.org/10.48550/arxiv.2510.23245 This dual representation supports both machine processing and human interpretability.A version control system tracks all policy modifications, ensuring a complete audit trail of how governance requirements evolve over time.... |
| [v3394] | Discovering Concept Directions from Diffusion-based Counterfactuals via Latent Clustering https://arxiv.org/abs/2505.07073 Among the various XAI paradigms, concept-based explanations have gained particular attention due to their ability to express model behavior in terms of high-level, semantically meaningful concepts, rather than low-level feature weights or pixel-base... |
| [v3396] | Trusted Data for AI Agents: Enterprise Foundation for Governance, Quality and Scale https://www.informatica.com/resources/articles/trusted-data-for-ai-agents-guide.html Regulatory requirements (GDPR, HIPAA, SOC 2) demand strict access controls, masking, lineage and auditability. In multi-agent systems, agent-specific accountability quickly becomes complicated without centralized governance. Governance by design. Ef... |
| [v3402] | BEM: Training-Free Background Embedding Memory for False-Positive Suppression in Real-Time Fixed-Background Camera https://arxiv.org/abs/2604.11714 BEM estimates clean background embeddings, maintains a prototype memory, and re-scores detection logits with an inverse-similarity, rank-weighted penalty, effectively reducing false positives while maintaining recall. Empirically, background-frame co... |
| [v3453] | Artificial Intelligence (AI) is becoming a crucial part of almost every industry. https://www.validaitor.com/post/understanding-the-basics-of-ai-testing Metamorphic and Property-Based Testing: AI systems often lack a clear test oracle (i.e., a known correct output). Metamorphic testing addresses this by checking whether the system behaves consistently under known transformations (e.g., image rotation... |
| [v3495] | Agentic AI pipelines are computational architectures where multiple specialized AI agents collaborate to complete complex tasks. https://www.exxactcorp.com/blog/deep-learning/agentic-ai-platforms-hardware-infrastructure Agentic AI pipelines are computational architectures where multiple specialized AI agents collaborate to complete complex tasks. ... This architecture is governed by a set of key principles designed to ensure scalability, security, and manageability:... |
| [v3561] | Secure Control of Connected and Automated Vehicles Using Trust-Aware Robust Event-Triggered Control Barrier Functions https://doi.org/10.14722/vehiclesec.2024.23037 Secure Control of Connected and Automated Vehicles Using Trust-Aware Robust Event-Triggered Control Barrier Functions --- 8} within the time interval [t i,k , t i,k+1 ) renders the set Ci and therefore C i forward invariant for the dynamic system def... |
| [v3577] | On Minimizing Adversarial Counterfactual Error in Adversarial Reinforcement Learning https://arxiv.org/abs/2406.04724 Deep Reinforcement Learning (DRL) policies are highly susceptible to adversarial noise in observations, which poses significant risks in safety-critical scenarios. The challenge inherent to adversarial perturbations is that by altering the informatio... |
| [v3604] | Efficient LLM Safety Evaluation through Multi-Agent Debate https://arxiv.org/abs/2511.06396 Sensitivity to rubric design, prompting context, and model-specific inductive biases yields poor inter-judge reliability and complicates alignment with human values, especially under semantic and adversarial conditions .These observations motivate ou... |
| [v3635] | Responsible AI in Customer Service: Guidelines https://customerscience.com.au/customer-experience-2/responsible-ai-customer-service-guidelines/ A purpose-built option is brand-aligned communication quality scoring with CommScore.AI. NIST. AI RMF Generative AI Profile. NIST AI 600-1, 2024. https://nvlpubs.nist.gov/nistpubs/ai/NIST.AI.600-1.pdf ISO/IEC. ISO/IEC 42001:2023 Artificial intellig... |
| [v3666] | Sparsity-Aware Unlearning for Large Language Models https://doi.org/10.48550/arXiv.2602.00577 However, existing methods are designed for dense models and overlook model sparsification-an essential technique for efficient LLM deployment. We find that unlearning effectiveness degrades substantially on sparse models. Through empirical analysis, ... |
| [v3671] | Multi-Abstractive Neural Controller: An Efficient Hierarchical Control Architecture for Interactive Driving https://doi.org/10.1109/lra.2023.3273421 We train this neural controller with real-world driving data via behavior cloning and show improved explainability, sample efficiency, and similarity to human driving. I. INTRODUCTION With robotic and autonomous driving applications expanding from ... |
| [v3855] | Greetings and welcome to the third edition of "Weekly AI News"! https://newsletter.chatwhisperer.ai/p/weekly-ai-news-110225 OpenAI now offers European data residency, helping local organisations comply with GDPR, Germany's Federal Data Protection Act, and other privacy regulations. Eligible API endpoints, plus new ChatGPT Enterprise and Edu accounts, can store data at res... |
| [v3946] | System and method for privately hosting machine learning models and collaborative computations https://patents.google.com/?oq=18899444 ... run, by the encrypted file system, a hardware attestation report comprising a cryptographically signed statement validating that the model host is running on a genuine processor manufactured by an enclave manufacturer with a secure compute elemen... |
| [v3950] | Spindle supports trust-weighted defeasible reasoning, enabling source attribution, trust-weighted conclusions, partial defeat (diminishment), and multi-perspective evaluation. https://spindle-rust.anuna.io/guides/trust ... d flies (trust: 0.90) [agent:coder] Each conclusion shows: The provability symbol (+D, -D, +d, -d) The literal The trust degree in parentheses The contributing sources in brackets Without --trust, conclusions display in the standard format ... |
| [v4009] | STAR-Teaming: A Strategy-Response Multiplex Network Approach to Automated LLM Red Teaming https://arxiv.org/abs/2604.18976 In the following sections, we detail each part of the framework: Section 3.2 describes the MAS pipeline, Section 3.3 explains the construction of the multiplex network, and Section 3.4 outlines the probabilistic strategy sampling procedure. Multi Age... |
| [v4152] | Discover IIT Bombay's new Agentic AI Certificate and access the program through Great Learning to build practical AI agent development skills. https://www.mygreatlearning.com/blog/access-the-agentic-ai-certificate-course-on-great-learning/ Discover IIT Bombay's new Agentic AI Certificate and access the program through Great Learning to build practical AI agent development skills. ... Reinforcement learning and reward training Prompt optimisation using DSPy Best-of-N sampling and feed... |
| [v4162] | REMIX-FND: A Multi-Modal Domain-Invariant Framework with Adaptive Evidence Retrieval for Cross-Domain Fake News Detection https://doi.org/10.66261/817fqh85 In addition, Monte Carlo dropout is employed for uncertainty-conditioned evidence retrieval depth, a Dynamic Source Reliability Graph (DSRG) for temporally decaying source reliability, and a six-detector ensemble for AI-generated text detection. The ... |
| [v4238] | FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning https://arxiv.org/abs/2511.14715 The server performs the entire multi-dimensional reputation assessment Section III-B and dynamic thresholding III-C on these noisy updates. This introduces a clear privacy-utility trade-off: the server's scoring mechanism must now distinguish between... |
| [v4257] | VectorSmuggle: Steganographic Exfiltration in Embedding Stores and a Cryptographic Provenance Defense https://arxiv.org/abs/2605.13764 VectorSmuggle: Steganographic Exfiltration in Embedding Stores and a Cryptographic Provenance Defense --- Abstract: Modern retrieval-augmented generation (RAG) systems convert sensitive content into high-dimensional embeddings and store them in vecto... |
| [v4260] | Beyond Black-Box Explanations: Monte Carlo Dropout for Uncertainty-Aware Explainable AI in Marketing Analytics https://doi.org/10.1109/EECSI67060.2025.11290147 Marketing AI systems increasingly rely on explainable artificial intelligence (XAI) to justify customer targeting, yet current methods provide no indication of when explanations can be trusted, creating risks of unreliable targeting and reduced campa... |
| [v4266] | Fugu-MT 論文翻訳(概要): When and Where to Attack? https://fugumt.com/fugumt/paper_check/2602.04356v1 Language Models)に対する敵対的攻撃は 現代のマルチモーダルシステムにおける安全性の脆弱性を明らかにするために重要である。 ランダムトリミングのような入力変換に基づく最近の攻撃は 空間的局所的な摂動は 大域的な画像操作よりも効果的であることを示唆している。 しかし 画像全体をランダムにトリミングすることは本質的に確率的であり ピクセルごとの摂動予算を効率的に使うことができない。 私たちは2つの重要な観察をします。 (i)地域注意スコアは 対向的損失感度と正の相関関係にあり (II)... |
| [v4281] | Quick Recap: Embeddings (vectors) are numerical representations of meaning. "" https://newsletter.aitechhive.com/p/vectorization-and-enterprise-indexing-theory Fail: <85% overlap indicates model is missing cases or including wrong ones By 2026, all financial institutions will run these validation tests quarterly. Embeddings that fail are retrained or replaced. Regulatory and Practical Context How Regulat... |
| [v4285] | LLM-assisted Agentic Edge Intelligence Framework https://arxiv.org/abs/2604.09607 To enhance system robustness and security, a dedicated component is introduced to validate faulty business logic, developed by LLMs, before further processing. 3. Our proposed framework is adaptive in nature, which generates lightweight code and cons... |
| [v4426] | Robust Explainable AI via Adversarial Latent Diffusion Models: Mitigating Gradient Obfuscation with Interpretable Feature Attribution https://doi.org/10.52783/jisem.v10i36s.6522 For explanation generation, Integrated Gradients was employed to produce interpretable feature attributions. The models were evaluated based on adversarial robustness, explanation stability (measured by Structural Similarity Index Measure, SSIM), and... |
| [v4465] | When to Re-embed Documents in Your Vector Database https://particula.tech/blog/when-to-reembed-documents-vector-database The most common reason to re-embed is switching to a more capable embedding model. If you initially implemented RAG with text-embedding-ada-002 and now want to use text-embedding-3-large, you need to re-embed all existing documents. Mixing embeddings... |
| [v4527] | Counterfactual Visual Explanation via Causally-Guided Adversarial Steering https://doi.org/10.48550/arXiv.2507.09881 To the best of our knowledge, no existing method well addresses these challenges, underscoring the need for a new approach that incorporates causal reasoning into the generation of counterfactual visual explanations. To address the aforementioned cha... |
| [v4568] | Medium Voltage Direct Current Shipboard Power Network Reconfiguration Using Graph-Based Reinforcement Learning https://doi.org/10.1115/1.4069035 The RL policy network is designed using a graph convolutional network (GCN). This technique optimizes the optimal status (ON/OFF) of switches in the MVDC shipboard power network, ensuring maximum power availability to loads during disruptive events s... |
| [v4581] | Agentic Artificial Intelligence (AI) Orchestration And Memory Systems Market to Reach $37.11B by 2030 at 40.2% CAGR https://www.einpresswire.com/article/909620759/agentic-artificial-intelligence-ai-orchestration-and-memory-systems-market-to-reach-37-11b-by-2030-at-40-2-cagr The agentic artificial intelligence (AI) orchestration and memory systems market is segmented by solution type into orchestration frameworks, memory layers or vector databases (DBs), workflow engines, context-management software development kits (SDK... |
| [v4628] | Understanding disentangling in β-VAE https://arxiv.org/abs/1804.03599 It is a modification of the Variational Autoencoder (VAE) objective, a generative approach that aims to learn the joint distribution of images x and their latent generative factors z. β-VAE adds an extra hyperparameter β to the VAE objective, which ... |
| [v4684] | Beyond Single-Point Judgment: Distribution Alignment for LLM-as-a-Judge https://doi.org/10.48550/arxiv.2505.12301 These results suggest that incorporating adversarial training enables the model to effectively align with all plausible distributions within the perturbation set, thereby improving robustness and fidelity in distributional alignment. Conclusion In ... |
| [v4783] | The Specialized High-Performance Network on Anton 3 - NewsBreak https://www.newsbreak.com/news/2491549896545/the-specialized-high-performance-network-on-anton-3 URL: Backdoor Defense with Machine Unlearning - https://newsbreak.com/news/2494719563784/backdoor-defense-with-machine-unlearning URL: Automated machine learning for secure key rate in discrete-modulated continuous-variable quantum key distribution ... |
| [v4801] | Mechanistic understanding and validation of large AI models with SemanticLens https://doi.org/10.1038/s42256-025-01084-w 'Auditing concept alignment with expected reasoning' describes how these functionalities provide the basis for effectively auditing the alignment of the reasoning of the model with respect to human expectation. We demonstrate how to spot flaws in med... |
| [v4846] | HyperTrust-Fog: Hypergraph-Based Trust-Aware-Federated Orchestration with Energy Adaptive Scheduling for Hierarchical Cloud Fog Edge Systems https://doi.org/10.21203/rs.3.rs-8230509/v1 It begins from the observation that many existing federated learning (FL) or graph-based orchestration methods rely on pairwise interaction models and largely static trust assumptions. Such systems are inadequate for fog environments where collaborat... |
| [v4851] | A multi-label visualisation approach for malware behaviour analysis https://doi.org/10.1038/s41598-025-21848-z To improve attribution reliability, we extend Gradient-weighted Class Activation Mapping (Grad-CAM) with a Bayesian formulation, enabling uncertainty-aware visualisation of discriminative regions linked to multiple categories. The regions identified ... |
| [v4896] | Introducing Dataset Q&A: Expanding natural language querying for structured datasets in Amazon Quick https://aws.amazon.com/blogs/machine-learning/introducing-dataset-qa-expanding-natural-language-querying-for-structured-datasets-in-amazon-quick/ Users can explore any dataset directly, going beyond what an author has pre-configured, while all the security, permissions, and governance that enterprises expect from Quick remain fully enforced. While the industry has raced to ship text-to-SQL de... |
| [v4930] | Actual costs may vary based on tokenization and usage patterns. https://calculatequick.com/ai/claude-token-cost-calculator/ Opus 4.5 introduces fine-grained control over reasoning depth. The effort parameter lets you balance performance versus cost on each API request. Low Effort Fastest responses with minimal reasoning depth. Best for simple tasks, quick classification... |
| [v4945] | How Much Does It Cost to Make A Crypto Wallet App on Blockchain? https://appinventiv.com/blog/ai-software-development-uae/ Filtering or masking sensitive fields before model access Security Is Built Into the Architecture AI introduces new risk surfaces, from prompt inputs to downstream integrations. In AI-powered software development in Dubai, security is not treated a... |
| [v4973] | System And Method For Website Analysis Using Computer Vision https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260120500).pn The system demonstrates improved performance characteristics compared to traditional DOM-based web scraping approaches. In empirical testing across diverse website types, the visual analysis approach maintained consistent extraction accuracy despite ... |
| [v5000] | Deep learning emerges as key shield for smart grid cybersecurity | Technology https://www.devdiscourse.com/article/technology/3340328-deep-learning-emerges-as-key-shield-for-smart-grid-cybersecurity However, FL itself introduces communication overhead and is still susceptible to poisoning attacks, where malicious nodes feed deceptive data into the learning process. Legacy system compatibility is another roadblock. Many current grid systems were... |
| [v5002] | In this paper, we focus on applications in machine learning, optimization, and control that call for the resilient selection of a few elements, e.g. features, sensors, or leaders, against a number of https://core.ac.uk/search/ In general, such resilient optimization problems are hard, and cannot be solved exactly in polynomial time, even though they often involve objective functions that are monotone and submodular.... |
| [v5037] | Beyond Binary Opinions: A Deep Reinforcement Learning-Based Approach to Uncertainty-Aware Competitive Influence Maximization https://doi.org/10.48550/arxiv.2504.15131 The belief ( bi ) and disbelief ( di ) are then recalibrated by subtracting their respective contributions to uncertainty, maintaining the overall probability distribution. We incorporate this UM in designing uncertainty-aware exploration-exploitati... |
| [v5041] | Why Сurrent LLMs Struggle to Integrate with Complex Data Lakes in Multi-agent Systems https://techbullion.com/why-%D1%81urrent-llms-struggle-to-integrate-with-complex-data-lakes-in-multi-agent-systems/ Column-based security restricts access to sensitive fields. Policy Awareness. LLMs lack an inherent understanding of column-level permissions and may retrieve restricted columns from LLM Chat Memory without guardrails. Metadata Exploitation . Attac... |
| [v5061] | Orchestrator-Agent Trust: A Modular Agentic AI Visual Classification System with Trust-Aware Orchestration and RAG-Based Reasoning https://doi.org/10.48550/arXiv.2507.10571 In summary, our contributions are fourfold: (1) A modular agentic AI system that decouples perception, reasoning, and retrieval; (2) a novel trust-aware orchestration strategy grounded in multidimensional calibration; (3) a CLIP-RAG-based re-evaluati... |
| [v5065] | RevenueGrid Blog All resources AI Readiness Checklist for FinServ: Are You Ready for AI Adoption? https://revenuegrid.com/blog/ai-readiness-checklist-finserv/ Automated PII detection runs before an LLM processes any data; masking or tokenization is applied by default. Role-based access control enforces least-privilege access for both users and AI assistants. Model Risk Classification Tiered model invento... |
| [v5088] | Explanation of Dynamic Physical Field Predictions using WassersteinGrad: Application to Autoregressive Weather Forecasting https://arxiv.org/abs/2604.22580 It is also interesting to remark that gradient-based techniques such as SmoothGrad are now standard on images to robustify the explanations using pointwise averages of the attribution maps obtained from several noised inputs. Our goal is to efficient... |
| [v5150] | Following our successful HULA framework workshops, we evolved the concept at Founders & Coders to explore a different challenge: how do development teams coordinate when each developer has their own https://www.maxitect.blog/posts/beyond-solo-ai-how-pair-programming-with-claude-code-transforms-team-development Teams following this approach progressed smoothly through feature development whilst those attempting full AI delegation found themselves rebuilding foundations as teammates moved ahead. Why live documentation tumps individual context The TICKETS.m... |
| [v5187] | Matrix Control Barrier Functions https://arxiv.org/abs/2508.11795 Matrix Control Barrier Functions --- a method increasingly used in robotics in fields such as SLAM, pose graph optimization, and sensor fusion. One recent work has begun to explore how control barrier functions can be used to ensure NLS remains well... |
| [v5212] | The Student Seminar Series is a student-operated platform where graduate students can present their research to their peers and practice their presentation skills and faculty have an opportunity to https://uwaterloo.ca/statistics-and-actuarial-science/student-seminar-series The Student Seminar Series is a student-operated platform where graduate students can present their research to their peers and practice their presentation skills and faculty have an opportunity to present their research to a student audience. ... Ph... |
| [v5233] | Batch reinforcement learning, also called offline reinforcement learning, is the process of training an RL policy using a fixed dataset of interactions collected beforehand, without further environme https://www.shadecoder.com/topics/batch-reinforcement-learning-a-comprehensive-guide-for-2025 When possible, integrate explainability and logging to trace policy decisions back to data. Overall, the process is iterative: success depends on data quality, conservative design, and disciplined offline validation. Common Mistakes with Batch Reinf... |
| [v5245] | Assessing the Impact of Requirement Ambiguity on LLM-based Function-Level Code Generation https://arxiv.org/abs/2604.21505 Even state-of-the-art models, such as GPT-4, exhibit a performance drop exceeding 30% when confronted with ambiguous specifications, suggesting that current benchmarks significantly overestimate the effectiveness of LLMs in real-world, "noisy" softwa... |
| [v5355] | TriGuard: Testing Model Safety with Attribution Entropy, Verification, and Drift https://doi.org/10.48550/arxiv.2506.14217 TriGuard draws upon and extends foundational research across adversarial robustness, formal verification, and interpretability.Our contribution lies in unifying these efforts under a shared evaluation framework and proposing a novel metric -Attributi... |
| [v5422] | Multi-Modal Fact-Verification Framework for Reducing Hallucinations in Large Language Models https://doi.org/10.48550/arXiv.2510.22751 This hallucination problem has become a major barrier to deploying these models in real-world applications where accuracy matters. We developed a fact verification framework that catches and corrects these errors in real-time by cross checking LLM ou... |
| [v5423] | Visual Disentangled Diffusion Autoencoders: Scalable Counterfactual Generation for Foundation Models https://doi.org/10.48550/arXiv.2601.21851 The oracle O is just another classifier that we distill the decision strategy of our original classifier f into. Because we train O from scratch, this avoids the weight-specific adversarial attacks that fool f also fool O. Gain To quantify the effect... |
| [v5472] | When outcomes carry risk-legal exposure, investment loss, or reputational damage-'good enough' AI isn't good enough. https://suprmind.ai/hub/insights/autonomous-ai-agents-a-practitioners-guide-to-multi-llm/ This includes user preferences, domain knowledge, and patterns learned from previous interactions. Context Fabric maintains this persistent context without requiring you to manually track conversation history. The challenge is managing context windo... |
| [v5481] | For AI safety researchers: Focus on Section II. https://aliveness.kunnas.com/articles/privilege-separation-ai-safety Adversarial dynamic: Research on Chain of Thought Monitorability (Korbak et al. 2024) finds this approach "fragile" - models hide reasoning when optimization pressure favors it. Timeline mismatch: Scalable mechanistic interpretability estimated at 1... |
| [v5523] | Predicting the epidemiological trend of acute hemorrhagic conjunctivitis in China using Bayesian structural time-series model https://doi.org/10.1038/s41598-024-68624-z The Bayesian Time Structure Sequence (BSTS), on the other hand, is a dynamic regression model that allows parameters to evolve over time, accurately capturing the random behavior of time series.This approach allows for variance control and the imposi... |
| [v5532] | Structure suggests 10040.5ImportanceReferenceImportance: 40.5/100How central this topic is to AI safety. https://www.longtermwiki.com/wiki/E174 The suite combines SAEs and transcoders to enable analysis of complex multi-step behaviors including jailbreaks, refusal mechanisms, and chain-of-thought faithfulness. Quantitative Progress Metrics Quantitative progress has accelerated dramatically... |
| [v5546] | Artificial intelligence agents in healthcare research: A scoping review https://doi.org/10.1371/journal.pone.0342182 The COVID-19 pandemic catalyzed the adoption of remote care modalities, creating an urgent need for digital tools capable of sustaining patient engagement and clinical continuity without physical contact .Concurrently, the maturation of large languag... |
| [v5547] | Amortized Latent Steering: Low-Cost Alternative to Test-Time Optimization https://doi.org/10.48550/arXiv.2509.18116 Test-time optimization remains impractical at scale due to prohibitive inference costs\textemdash techniques like iterative refinement and multi-step verification can require $10$--$100\times$ more compute per query than standard decoding. Latent spa... |
| [v5569] | RAIN: Secure and Robust Aggregation under Shuffle Model of Differential Privacy https://arxiv.org/abs/2603.03108 Secure aggregation is a foundational building block of privacy-preserving learning, yet achieving robustness under adversarial behavior remains challenging. ... Overall, these results indicate that signspace representation effectively lowers client-s... |
| [v5583] | The pervasive influence of recommender systems across digital landscapes necessitates continuous innovation to overcome inherent limitations and enhance user experience. https://creativenews.io/research-reports/advancements-in-social-trust-integration-for-recommender-systems-a-comprehensive-review/ Recommendations are then generated by aggregating ratings from trusted users, weighted by this propagated trust score. MoleTrust (Massa & Avesani, 2007): Similar to TidalTrust, MoleTrust also considers trust propagation but emphasizes the local prop... |
| [v5586] | Tiny-Critic RAG: Empowering Agentic Fallback with Parameter-Efficient Small Language Models https://doi.org/10.48550/arxiv.2603.00846 Retrieval-Augmented Generation (RAG) grounds Large Language Models (LLMs) to mitigate factual hallucinations. ... RAGAS Faithfulness.b CPQ: Explicit routing Cost Per 10k Queries in USD.c CPQ estimations assume an average context of 2K tokens under op... |
| [v5599] | Traditional reinforcement learning-based robotic control methods are often task-specific and fail to generalize across diverse environments or unseen objects and instructions. https://aclanthology.org/people/deepanway-ghosal/unverified/ In this work, we propose the Embodied Multimodal Action Model with Grounded Chain of Thought and Look-ahead Spatial Reasoning, EMMA-X. EMMA-X leverages our constructed hierarchical embodiment dataset based on BridgeV2, containing 60,000 robot manipul... |
| [v5635] | SCI-IoT: A Quantitative Framework for Trust Scoring and Certification of IoT Devices https://arxiv.org/abs/2511.18045 12]. The following section outlines the major vulnerability classes, associated real world incidents, and the corresponding mitigation expectations aligned with Grades A-F of the proposed certification framework. Insecure Communication Protocols L... |
| [v5668] | RzkFL: a Verifiable, Fast and Privacy-Preserving Framework for Federated Learning Inference Using Recursive Zero-Knowledge Proofs and on-Chain Verification https://doi.org/10.1109/blockchain67634.2025.00028 RzkFL: a Verifiable, Fast and Privacy-Preserving Framework for Federated Learning Inference Using Recursive Zero-Knowledge Proofs and on-Chain Verification... |
| [v5695] | Goodhart's Law Applies to NLP's Explanation Benchmarks https://doi.org/10.18653/v1/2024.findings-eacl.88 Slack et al. demonstrate how one could exploit the OOD issue to manipulate the feature importance ranking from LIME and SHAP and conceal problems vis-a-vis fairness.They propose an adversarial wrapper classifier designed such that a sensitive featur... |
| [v5720] | FedRio: Personalized Federated Social Bot Detection via Cooperative Reinforced Contrastive Adversarial Distillation https://arxiv.org/abs/2604.10678 We first introduce an adaptive message-passing module as the graph neural network backbone for each client. To facilitate efficient knowledge sharing of global data distributions, we design a federated knowledge extraction mechanism based on generati... |
| [v5732] | PolySwarm: A Multi-Agent Large Language Model Framework for Prediction Market Trading and Latency Arbitrage https://arxiv.org/abs/2604.03888 PolySwarm system design and implementation: a production-ready multi-agent LLM trading terminal deploying 50 diverse personas on Polymarket with full architectural description, asynchronous execution pipeline, and paper/live trading modes. Confidence... |
| [v5769] | MSDA-GDS: A Dual-Branch Hybrid Federated Explainable Deep Learning Framework for CAN Bus Intrusion Detection in Internet of Vehicles https://doi.org/10.19139/soic-2310-5070-3599 The framework integrates Apache Spark-accelerated preprocessing, FedProx federated learning with differential privacy, and multi-method explainability (SHAP, LIME, gradient saliency).... |
| [v5815] | Use the AI STAR Method Generator to produce structured behavioral interview diagrams in seconds. https://creately.com/diagram/example/3KKZufKnFz8/ai-star-interview-method-template Generate audit-ready reports, trace decision rationale, and maintain secure logs to meet GDPR and SOC 2 Type 2 requirements.... |
| [v5831] | Generative artificial intelligence in diabetes healthcare https://doi.org/10.1016/j.isci.2025.113051 This can be achieved by enforcing temporal ordering, integrating structural causal models, or training on interventional and counterfactual data. In this context, graph-based techniques such as Graph Neural Networks (GNNs) provide powerful tools for ... |
| [v5920] | A Framework for Modeling Cognitive Processes in Intelligent Agents Using Behavior Trees https://doi.org/10.1145/3749566.3749619 In this way, we use an exploration technique based on pairing a combined behavior tree with the target model. We empirically show that our framework is effective in four benchmark MARL domains. Moreover, the results of a user study show that the gene... |
| [v6008] | SoK: Security of Autonomous LLM Agents in Agentic Commerce https://arxiv.org/abs/2604.15367 A critical finding of our analysis is that the most dangerous attacks on autonomous financial agents exploit crosslayer interactions, where a vulnerability at one layer triggers a cascading failure at another.We identify and characterize all 12 cross... |
| [v6031] | MedMMV: A Controllable Multimodal Multi-Agent Framework for Reliable and Verifiable Clinical Reasoning https://doi.org/10.48550/arXiv.2509.24314 By controlling instability through a verifiable, multi-agent process, our framework provides a robust path toward deploying trustworthy AI systems in high-stakes domains like clinical decision support.... |
| [v6049] | AW-GATCN: Adaptive Weighted Graph Attention Convolutional Network for Event Camera Data Joint Denoising and Object Recognition https://doi.org/10.1109/IJCNN64981.2025.11227212 For noise reduction, inspired by , we employ an adaptive algorithm that dynamically adjusts the weighting radius based on multiple event point features, filtering out noise. These weights are then integrated with a graph attention mechanism to select... |
| [v6164] | Emerging multi-robot systems rely on cooperation between humans and robots, with robots following automatically generated motion plans to service application-level tasks. https://doi.org/10.48550/arxiv.2301.10704 Distributed resilient submodular action selection in adversarial environments. IEEE Robotics and Automation Letters 6, 3 (2021), 5832-5839. [Morante et al.(2015)] Santiago Morante, Juan G Victores, and Carlos Balaguer. 2015. Cryptobotics: Why robots ... |
| [v6171] | What does it mean to connect unstructured data in a vector database to an LLM in a RAG pipeline? https://airbyte.com/data-engineering-resources/connecting-vector-database-to-llm-in-rag-pipeline Align them with your corpus and serving constraints. Retrieval tactics: similarity search vs hybrid approaches Vector similarity search finds semantically close chunks from embeddings. Hybrid retrieval combines semantic vectors with lexical methods... |
| [v6219] | この記事を一言で要約すると 反実仮想的な説明に基づく機械学習モデル解釈手法に対する Microsoft Research の取り組みと その成果 (アルゴリズム) を८ https://qiita.com/OpenJNY/items/ef885c357b4e0a1551c0 Support for other algorithms for generating counterfactual explanations Incorporating causal constraints when generating counterfactual explanations 機械学習モデルの解釈手法も成熟してきつつあり 原先生 の Lasso 解列挙手法 [AAAI 2017] のような 解釈した先の意識決定を意識するフェーズに来ているのかなと思いました。 そのような... |
| [v6223] | Method and apparatus for combining data to construct a floor plan https://patents.google.com/?oq=17876634 The gradient ∇ƒ(x) of the function ƒ(x) may be a vector including all first partial derivatives. The matrix including all first partial derivatives may be the Jacobian while the matrix including all the second derivatives may be the Hessian, (2023)... |
| [v6236] | Explaining Hypergraph Neural Networks: From Local Explanations to Global Concepts https://doi.org/10.48550/arXiv.2410.07764 The implanted motifs reflect human reasoning, but are not necessarily faithful to the neural network, which may instead rely on a variant or correlate of the motif. Rather, a good explanation should provide users information about the hyperGNN's pred... |
| [v6260] | GitHub - tigerneil/awesome-deep-rl: For deep RL and the future of AI. https://github.com/tigerneil/awesome-deep-rl Language as an Abstraction for Hierarchical Deep Reinforcement Learning 18 Jun 2019 arxiv Variational Option Discovery Algorithms 26 July 2018 A Laplacian Framework for Option Discovery in Reinforcement Learning 16 Jun 2017 Robust Imitation of Div... |
| [v6270] | Gaussian Amplitude Amplification for Quantum Pathfinding https://pubmed.ncbi.nlm.nih.gov/35885186/ We study an oracle operation, along with its circuit design, which combined with the Grover diffusion operator boosts the probability of finding the minimum or maximum solutions on a weighted directed graph. We focus on the geometry of sequentially c... |
| [v6280] | A take on a new threat from an old adversaryYou're already thinking about compliance - is digital accessibility on your list? https://www.packtpub.com/en-cy/newsletters/secpro The post is frequently cited in operator and VC circles for its market intelligence and strategic forecasting.This week's academiaFederated Learning-Driven Cybersecurity Framework for IoT Networks with Privacy-Preserving and Real-Time Threat Detectio... |
| [v6294] | Recourse provides individuals who received undesirable labels (e.g., denied a loan) from algorithmic decision-making systems with a minimum-cost improvement suggestion to achieve the desired outcome. https://arxiv.org/html/2509.21293v1 In particular, we measure model changes by bounding the LpL^{p} norm of the difference between initial and changed models, where p ≥ 1p\geq 1 but p≠∞p\neq\infty. We provide a new algorithm that provably computes the optimal robust recourse for genera... |
| [v6300] | Detecting Concept Drift with SHapley Additive ExPlanations for Intelligent Model Retraining in Energy Generation Forecasting https://doi.org/10.1007/978-3-032-08324-1_7 Detecting Concept Drift with SHapley Additive ExPlanations for Intelligent Model Retraining in Energy Generation Forecasting --- This study introduces a novel approach that leverages SHapley Additive Explanations (SHAP) to dynamically detect concept ... |
| [v6331] | Conduction and entropy analysis of a mixed memristor-resistor model for neuromorphic networks https://doi.org/10.1088/2634-4386/acd6b3 Thus, network entropy is used to understand the self-reinforcing and cooperative inhibition of other memristive elements resulting in the formation of a winner-take-all path. Both the low interaction strength and the dilution of the memristive fracti... |
| [v6337] | With the increasing integration of a high proportion of renewable energy, the fluctuation characteristics of distributed power generation such as wind and photovoltaic energy affect the safe and stab https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2025.1416309/full A novel metric to quantify and enable resilient distribution system using graph theory and Choquet integral. Smart Grid9 (4), 2918 - 2929. 2016.2623818 SeivastavaA. K. (2016). Defining and enabling resiliency of electric distribution systems with mu... |
| [v6371] | Human-Centered LLM-Agent System for Detecting Anomalous Digital Asset Transactions https://arxiv.org/abs/2510.20102 Large-Scale User Validation: Conduct IRB-approved studies to generalize trust and interpretability findings. Conclusion The accelerating complexity of digital asset ecosystems demands anomaly detection systems that are not only technically advanced... |
| [v6398] | Resource-Efficient Medical Image Classification for Edge Devices https://doi.org/10.1109/icamida64673.2025.11209605 An emerging solution to this challenge is Saliency Guided Training, which integrates interpretability into the training process.By iteratively masking less relevant input features-those with low gradients-and enforcing consistent outputs for masked a... |
| [v6422] | This guide analyzes Atlas, CLOiD, Spirit v1.5 benchmarks, tools, and predictions. https://globzette.com/technology/embodied-ai-beyond-the-chatbot-2026/ This guide analyzes Atlas, CLOiD, Spirit v1.5 benchmarks, tools, and predictions. Move from research pilots to factory/home deployment with proven strategies. ... Open-source tactile/multi-agent reasoning excels. Production-ready for warehouses/facto... |
| [v6460] | Conformal Feedback Alignment: Quantifying Answer-Level Reliability for Robust LLM Alignment https://arxiv.org/abs/2601.17329 Ang Li, Qiugen Xiao, Peng Cao, Jian Tang, Yi Yuan, Zijie Zhao, Xiaoyuan Chen, Liang Zhang, Xiangyang Li, arXiv:2403.083092024arXiv preprintKaitong Yang, and 1 others Generating with confidence: Uncertainty quantification for black-box large language... |
| [v6569] | On the Hardness of Decentralized Multi-Agent Policy Evaluation under Byzantine Attacks https://doi.org/10.48550/arxiv.2409.12882 3) Main theoretical results: The following theorems state that, in the presence of Byzantine agents, no algorithm ensures that the normal agents' parameters converge to a fixed point in Problem 2. Theorem 1.When f > 0, Problem 2 is not solvable.Theo... |
| [v6706] | Explainability-Based Token Replacement on LLM-Generated Text https://doi.org/10.48550/arXiv.2506.04050 Beyond SHAP and LIME, alternative explainability approaches have been explored for NLP tasks. SyntaxShap extends SHAP by incorporating syntactic structure, assigning importance scores to phrase-level constituents rather than individual tokens, which... |
| [v6719] | An Explainable AI Framework for Image Analytics and Synthetic Image Creation Using CNN and GAN Architectures https://doi.org/10.14445/23488387/ijcse-v13i2p101 The framework also presented model-level, feature-level, and instance-level interpretability of CNN classifiers through gradient-based attribution, concept activation vectors, and saliency-based analysis of attention. Meanwhile, explainability is inh... |
| [v6743] | Ferret, a new Multimodal Large Language Model, excels in spatial referring and grounding within images using a hybrid region representation, achieving superior performance in multimodal tasks and red https://huggingface.co/papers/2310.07704 Ferret, a new Multimodal Large Language Model, excels in spatial referring and grounding within images using a hybrid region representation, achieving superior performance in multimodal tasks and reducing object hallucination.... |
| [v6781] | Group Lasso Based Selection for High - Dimensional Mediation Analysis https://doi.org/10.1002/sim.70351 For each model, sample N times its parameters according to their multivariate sampling distribution, and obtain the vectors or parameters Θ Y (n) and Θ Z (n) = Θ 1 (n) , . . . , Θ Kmax(n) , for n = 1, . . ., N .As in , the law of the parameters is ... |
| [v6784] | As LLM-based agents increasingly operate in multi-agent systems, understanding adversarial manipulation becomes critical for defensive design. https://verso.uidaho.edu/esploro/outputs/preprint/Intentional-Deception-as-Controllable-Capability-in/996896856401851 As LLM-based agents increasingly operate in multi-agent systems, understanding adversarial manipulation becomes critical for defensive design. We present a systematic study of intentional deception as an engineered capability, using LLM-to-LLM intera... |
| [v6815] | Encrypted Spiking Neural Networks Based on Adaptive Differential Privacy Mechanism https://doi.org/10.3390/e27040333 Based on the correlation between the model's output and the labels, as well as the differential privacy parameters, an adaptive noise scale is dynamically determined.... |
| [v6849] | Towards a Cognitive Meta-Model for Adaptive Trust and Reputation in Open Multi-Agent Systems https://doi.org/10.65109/xpvb5485 In this paper, a cognitive meta-model for adaptive trust and reputation in open multi-agent systems is presented. It acts as a complement to a non-adaptive model by allowing the agent to reason about it and react to changes in the environment. We dem... |
| [v6901] | Generalized Multi-Relational Graph Convolution Network https://arxiv.org/abs/2006.07331 Most GCN methods are either restricted to graphs with a homogeneous type of edges (e.g., citation links only), or focusing on representation learning for nodes only instead of jointly optimizing the embeddings of both nodes and edges for target-drive... |
| [v6912] | Measuring the Fragility of Trust: Devising Credibility Index via Explanation Stability (CIES) for Business Decision Support Systems https://arxiv.org/abs/2603.05024 Research demonstrates that widely used post hoc methods such as LIME and SHAP can be manipulated: adversarial scaffolding can conceal underlying biases while generating seemingly benign explanations . Likewise, adversarial perturbations can produce i... |
| [v7024] | Detectability Thresholds for Network Attacks on Static Graphs and Temporal Networks: Information-Theoretic Limits and Nearly-Optimal Tests https://arxiv.org/abs/2509.10925 We quantify how thresholds deform under bounded perturbations of the edge set (e.g., a small adversarial rewiring budget) and under mild model misspecification (e.g., modest heterogeneity in baseline edge probabilities or intensity drift).In our anal... |
| [v7032] | System and method for automated affinity-based network expansion through intelligent relationship discovery and compatibility matching https://patents.google.com/?oq=19298256 The method of claim 10, wherein the method further comprises the steps of: calculating affinity-based user acquisition coefficients in real-time using cohort analysis to measure exponential growth effectiveness; implementing propagation pathway opt... |
| [v7040] | Multi-Domain Adversarial Variational Bayesian Inference for Domain Generalization https://doi.org/10.1109/tcsvt.2022.3232112 Multi-Domain Adversarial Variational Bayesian Inference for Domain Generalization... |
| [v7081] | DSSA-TCN: Exploiting adaptive sparse attention and diffusion graph convolutions in temporal convolutional networks for traffic flow forecasting https://doi.org/10.1371/journal.pone.0336787 As shown in Fig 1, the model first transforms the raw inputs into a latent representation through a linear projection, and augments it with time-of-day, day-of-week, and learnable node embeddings. These embeddings help the model capture periodic traf... |
| [v7092] | MotionLM: Multi-Agent Motion Forecasting as Language Modeling https://doi.org/10.48550/arxiv.2309.16534 Of the existing joint prediction approaches, some apply a separation between marginal trajectory generation and interactive scoring .For example, Luo et al. initially produce a small set of marginal trajectories for each agent independently, before ... |
| [v7122] | Complex networks in Air Force-relevant applications, including multi-vehicle control, energy systems, and neuronal networks, are expected to guarantee performance, stability, and availability. https://hydra.ece.uw.edu/index.html At present, there is no computationally tractable analytical framework for modeling and designing resilient networks with provable performance guarantees. We propose to research and develop a submodular optimization framework for resilient complex n... |
| [v7128] | Offline-to-Online Multi-Agent Reinforcement Learning with Offline Value Function Memory and Sequential Exploration https://doi.org/10.65109/whoy8671 This improves online learning efficiency, as the offline pre-trained policy can focus on targeted exploration rather than an exhaustive random search of the action space, which is typically required when training from scratch. Offline MARL.The princ... |
| [v7130] | When Large Language Models Meet Personalization: Perspectives of Challenges and Opportunities https://doi.org/10.48550/arxiv.2307.16376 In each dialogue turn, the system needs to decide whether to ask the user a question or provide a recommendation. The decision-making process, particularly regarding which attribute to ask about, is typically handled by a policy network. On the other... |
| [v7136] | FedJudge: Blockchain-based full-lifecycle trustworthy federated learning incentive mechanism https://doi.org/10.1109/trustcom60117.2023.00066 This implementation guarantees a trustworthy incentive mechanism throughout the federated learning process. Through empirical validation and analysis on authentic datasets, we demonstrate that FedJudge significantly enhances Byzantine fault tolerance... |
| [v7214] | AI safetyBiosecurityCause prioritizationEffective givingExistential riskCareer choiceLong-Term Future FundEffective Altruism FundsLong-term futureThinking at the marginFunding opportunitiesGiving Sea https://forum.effectivealtruism.org/posts/qXWgFyQNgoijBzgwv/the-grant-decision-boundary-recent-cases-from-the-long-term This part-time project aims to create transparent, programmatic replacements for sparse autoencoder neurons in language models by developing symbolic representations in Python, evaluating their predictive accuracy, and measuring their impact on model... |
| [v7273] | Position: Introspective Experience from Conversational Environments as a Path to Better Learning https://arxiv.org/abs/2602.14910 When multi-agent systems are permitted to optimize their own communication protocols, they frequently converge on "Neuralese"-continuous vector-based exchanges that maximize information density and transmission speed.The LatentMAS framework recently ... |
| [v7283] | The internet has come a long way since its inception. https://smartechnews.com/featured/web-3-0-could-make-your-online-life-less-frustrating/ Web 3.0's transparent and tamper-evident nature will ensure that online interactions are more accountable than ever. With blockchain's immutable ledger, users can trust that their transactions and interactions are recorded accurately and transparentl... |
| [v7325] | Spatial Preference Rewarding for MLLMs Spatial Understanding https://doi.org/10.48550/arXiv.2510.14374 Compared to the baseline, SPR enhances MLLMs on both referring and grounding benchmarks, especially under higher IoU thresholds which demand higher localization accuracy. In addition, SPR can improve MLLM trustworthiness and reduce MLLM hallucination... |
| [v7329] | Adversarial robustness of amortized Bayesian inference https://doi.org/10.48550/arXiv.2305.14984 Here, we study the adversarial robustness of amortized Bayesian inference, focusing on simulation-based estimation of multi-dimensional posterior distributions. (2023)... |
| [v7366] | Proving a Photo Is Real Is Now Harder Than Faking ... https://www.albis.news/perspectives/proving-photos-real-harder-than-faking-them-2026 That's the idea behind C2PA - the Coalition for Content Provenance and Authenticity. It's an open standard backed by Adobe, Microsoft, Google, Intel, the BBC, and about 6,000 other organizations through the Content Authenticity Initiative. Instead of... |
| [v7389] | METR (where I work, though I'm cross-posting in a personal capacity) evaluated GPT-5 before it was externally deployed. https://www.lesswrong.com/posts/SuvWoLaGiNjPDcA7d/metr-s-evaluation-of-gpt-5 However, it remains unclear to what extent these performance gains can be attributed to human-like task decomposition or simply the greater computation that additional tokens allow. We show that transformers can use meaningless filler tokens (e.g., '... |
| [v7408] | As an awardee, Vasisht will receive a $25,000 USD stipend and the opportunity to intern with IBM to improve his understanding of industrial research, broaden his range of technical contacts, and str https://uwaterloo.ca/computer-science/news/vasisht-duddu-awarded-2024-ibm-phd-fellowship His approach uses machine learning, cryptographic techniques, and trusted hardware to enable companies to validate their claims. This work resulted in a paper titled Attesting Distributional Properties of Training Data for Machine Learning, presented... |
| [v7413] | In Part 4, we opened up the anatomy of an autonomous agent - the Intelligence Core that reasons over goals and the Trust Layer that governs what actions are permissible. https://www.wipro.com/engineering/articles/scaling-trust-in-autonomous-operations-with-agentic-ops-and-agentic-os/ Observability and Continuous Improvement: Agents generate structured reasoning logs, performance metrics, and decision traces. This observability layer allows engineers to audit agent conclusions, detect when model behaviour is drifting from expectat... |
| [v7414] | Learning Interaction-Aware Trajectory Predictions for Decentralized Multi-Robot Motion Planning in Dynamic Environments https://doi.org/10.1109/lra.2021.3061073 E. Decentralized Multi-Robot Motion Planning Having the trained trajectory prediction model, we can incorporate it with the MPC framework and solve the problem (2) in a decentralized manner. As shown in Fig. 1, in a multi-robot navigation scenario, ... |
| [v7423] | Faster search by lackadaisical quantum walk https://doi.org/10.1007/s11128-018-1840-y We perform a discrete-time coined quantum walk on this weighted graph while querying a Grover-type oracle that flips the sign of the amplitude at the marked vertex. (2018)... |
| [v7456] | Cyberlanguage: Native Communication for the Cyber-Physical-Social-Thinking Fusion Space https://arxiv.org/abs/2603.17498 Empirical development requires CyberCorpus: a multimodal interaction corpus annotated with four-dimensional labels (P, S, T, C components and their cross-dimensional mappings).Candidate data sources include human-robot task logs, smart-home interacti... |
| [v7542] | Optimizing Graph Causal Classification Models: Estimating Causal Effects and Addressing Confounders https://arxiv.org/abs/2602.17941 The intervention on a subset of nodes ⊆ modifies node features to produce an intervened graph ' with updated features ' : ' = (, ), where (.) denotes the controlled modification of node features for the intervened nodes.This enables to analyse how in... |
| [v7694] | A Novel Architectural Framework on IoT Ecosystem, Security Aspects and Mechanisms: A Comprehensive Survey https://doi.org/10.1109/ACCESS.2022.3207472 509 certificate that binds it to its authority name and is signed by a third party (trusted root). Nodes in this mode must support the same cipher suite as RPK mode. Moreover, in this mode, a node has also a list of trusted roots for certificate vali... |
| [v7702] | DNR: A Tunable Robust Pruning Framework Through Dynamic Network Rewiring of DNNs https://doi.org/10.1145/3394885.3431542 These trends suggest that our robustness is not achieved via gradient obfuscation . Generalized Robustness Against PGD Attack of Different Strengths CONCLUSIONS This paper addresses the open problem of achieving ultra-high compression of DNN model... |
| [v7725] | Process And System For Securely Searching And Summarizing Data From Source Systems https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260127209).pn provide the retrieved data and the correlated information to the operator. 2. The system of claim 1, wherein the one or more physical processors are further configured by the machine-readable instructions to dynamically generate harmonization steps ... |
| [v7814] | 6 proven lessons from the AI projects that broke before they scaled https://venturebeat.com/ai/6-proven-lessons-from-the-ai-projects-that-broke-before-they-scaled Prioritize explainability with tools like SHAP (SHapley Additive exPlanations) to build trust with stakeholders. Lesson 4: Ignoring deployment realities A model that shines in a Jupyter Notebook can crash in the real world. For example, a company's ... |
| [v7842] | Overcoming Data Loss in Wearable Disease Detection with GAN-Based Imputation https://doi.org/10.1038/s41746-026-02518-4 High rates of missing data in wearable sensor streams hinder early detection of infectious diseases, especially in low-resource settings with inconsistent device adherence and connectivity. We developed a lightweight generative adversarial network (G... |
| [v7928] | Static Sandboxes Are Inadequate: Modeling Societal Complexity Requires Open-Ended Co-Evolution in LLM-Based Multi-Agent Simulations https://doi.org/10.48550/arXiv.2510.13982 The development of genuinely open-ended, co-evolutionary simulations necessitates the concurrent evolution of agents and environments, fostering a continuous cycle of challenge and adaptation (Wang et al. 2023;Verma et al. 2023). Realization of this ... |
| [v7962] | Immutable Explainability: Fuzzy Logic and Blockchain for Verifiable Affective AI https://doi.org/10.48550/arXiv.2512.11065 Second, audit logs often lack reliability, as the entity operating the system may alter them. In this work, we introduce the concept of Immutable Explainability, an architecture designed to address both challenges simultaneously. Our approach combine... |
| [v7987] | Simplified Action Decoder for Deep Multi-Agent Reinforcement Learning https://www.emergentmind.com/papers/1912.02288 The SAD method incorporates best practices from recent advances in deep learning and reinforcement learning literature, such as recurrent neural networks to manage partial observability, distributed training frameworks improving sample efficiency, an... |
| [v8042] | Cooperative Observer-Based $\mathcal{H}_\infty$ Fault-Tolerant Tracking Control for Networked Processes with Sensor Faults https://arxiv.org/abs/2604.03921 Simulations on star, cyclic, and path topologies with heterogeneous agents confirm reliable tracking despite abrupt sensor faults and bounded disturbances, demonstrating a scalable and resilient coordination strategy for multi-agent systems with sens... |
| [v8051] | DRP: Distilled Reasoning Pruning with Skill-aware Step Decomposition for Efficient Large Reasoning Models https://arxiv.org/abs/2505.13975 Abstract: While Large Reasoning Models (LRMs) have demonstrated success in complex reasoning tasks through long chain-of-thought (CoT) reasoning, their inference often involves excessively verbose reasoning traces, resulting in substantial inefficien... |
| [v8072] | JAX-Privacy: A library for differentially private machine learning https://arxiv.org/abs/2602.17861 The library provides verified, modular primitives for critical components for all aspects of the mechanism design including batch selection, gradient clipping, noise addition, accounting, and auditing, and brings together a large body of recent resea... |
| [v8129] | Never Compromise to Vulnerabilities: A Comprehensive Survey on AI Governance https://arxiv.org/abs/2508.08789 For LLMs, alignment via RLHF provides foundational safety, but must be reinforced with runtime defenses such as input perplexity filters , circuit breakers , or ensemble-based rewriting frameworks like AutoDefense , MoGU .These defenses mitigate jail... |
| [v8175] | NeuroShield: A Neuro-Symbolic Framework for Adversarial Robustness https://arxiv.org/abs/2601.13162 We introduce \DesignII, a neuro-symbolic framework that integrates symbolic rule supervision into neural networks to enhance both adversarial robustness and explainability. Domain knowledge is encoded as logical constraints over appearance attributes... |
| [v8260] | Co-ordinated Tracking and Planning Using Air and Ground Vehicles https://doi.org/10.1007/978-3-642-00196-3_16 Similarly, the person is very small in the image, although relatively distinct; as a result, the motion of the helicopter makes the tracker lose track almost immediately without the ego-motion estimation. As a result, we use a motion model coupled w... |
| [v8265] | HalluScan: A Systematic Benchmark for Detecting and Mitigating Hallucinations in Instruction-Following LLMs https://arxiv.org/abs/2605.02443 We present HalluScan, a comprehensive benchmark framework that systematically evaluates hallucination detection and mitigation across 72 configurations spanning 6 detection methods, 4 open-weight model families, and 3 diverse domains. We introduce th... |
| [v8296] | Uncovering the non-equilibrium stationary properties in sparse Boolean networks - NewsBreak https://www.newsbreak.com/news/2515379035731/uncovering-the-non-equilibrium-stationary-properties-in-sparse-boolean-networks This is a form of test-time training that creates a self-supervised learning problem on test samples before performing the prediction task. In this way, our method enables efficient adaptation of encoded representations to evolving distributions, lea... |
| [v8322] | Automatic Document Editing for Improved RankingNiv Bardas, Tommy Mordo, Oren Kurland, Moshe Tennenholtz. https://researchr.org/alias/moshe-tennenholtz ... icdcs 2021: 954-964 Multi-issue social learningGal Bahar, Itai Arieli, Rann Smorodinsky, Moshe Tennenholtz. mss, 104:29-39, 2020. [ Fiduciary BanditsGal Bahar, Omer Ben-Porat, Kevin Leyton-Brown, Moshe Tennenholtz. icml 2020: 518-527 VCG under S... |
| [v8414] | Home Artificial Intelligence The Multi-Agent Trap | https://singularityfeed.com/the-multi-agent-trap-towards-data-science/ Unstructured multi-agent networks amplify errors as much as 17.2 instances in comparison with single-agent baselines. Not 17% worse. Seventeen instances worse. When brokers are thrown collectively with out structured topology (what the paper calls ... |
| [v8446] | Bayesian Dynamic Causal Discovery https://www.semanticscholar.org/paper/ec16fdb759d4a169d01905822be1e7d8ca885e85 Bayesian causal discovery methods tackle this problem by learning a posterior over the set of admissible graphs that are equally likely given our priors and observations. (2022)... |
| [v8447] | Posted on September 7, 2020 January 21, 2021 by Mike Gianfagna https://semiwiki.com/ip/dolphin-design/290385-dolphin-design-delivering-high-performance-audio-processing-with-tsmcs-22ull-process/ The figure below illustrates the high-performance and ultra-low power audio processing they can deliver for voice detection. The Dolphin approach for voice detection provides the following benefits: Stand-alone IP embedding a smart algorithm to det... |
| [v8492] | TRUST Agents: A Collaborative Multi-Agent Framework for Fake News Detection, Explainable Verification, and Logic-Aware Claim Reasoning https://arxiv.org/abs/2604.12184 Although supervised encoders remain stronger on raw metrics, TRUST Agents improves interpretability, evidence transparency, and reasoning over compound claims. Results also show that retrieval quality and uncertainty calibration remain the main bottl... |
| [v8528] | Stable Language Guidance for Vision-Language-Action Models https://arxiv.org/abs/2601.04052 Abstract: Vision-Language-Action (VLA) models have demonstrated impressive capabilities in generalized robotic control; however, they remain notoriously brittle to linguistic perturbations. We identify a critical ``modality collapse'' phenomenon wher... |
| [v8549] | WebGen-R1: Incentivizing Large Language Models to Generate Functional and Aesthetic Websites with Reinforcement Learning https://arxiv.org/abs/2604.20398 As shown in Figure 6, WebGen-R1 consistently outperforms a range of state-of-the-art proprietary and open-source baselines, such as DeepSeek-R1, GPT-5, and Qwen3-32B, on AAS. This suggests that WebGen-R1 has learned architecture-level and style-level... |
| [v8713] | Differential Privacy Integrated Federated Learning for Power Systems: An Explainability-Driven Approach https://doi.org/10.32604/cmc.2025.065978 Differential Privacy Integrated Federated Learning for Power Systems: An Explainability-Driven Approach... |
| [v8734] | Reinforcement Learning (RL) has emerged as a pivotal and transformative subset of machine learning, enabling autonomous agents to acquire optimal behaviors and decision-making policies through iterat https://medtechnews.uk/research-reports/reinforcement-learning-a-comprehensive-exploration-of-its-fundamentals-algorithms-historical-development-and-applications-across-industries/ However, the widespread and responsible deployment of RL systems hinges on diligently addressing several critical challenges. The inherent demand for vast amounts of interaction data necessitates ongoing research into sample-efficient learning, inclu... |
| [v8752] | A Unified Framework for Evaluating and Enhancing the Transparency of Explainable AI Methods via Perturbation-Gradient Consensus Attribution https://arxiv.org/abs/2412.03884 We propose Perturbation-Gradient Consensus Attribution (PGCA), a novel XAI method that fuses dense perturbation-based importance with Grad-CAM++ spatial precision through a five-stage pipeline comprising dual-strategy perturbation, gradient-based ref... |
| [v8781] | A comfortable graph structure for Grover walk https://doi.org/10.1088/1751-8121/acd735 The time evolution is determined by the Grover matrices assigned at each vertex: for each vertex u and each time step, the transmitting weight is 2/ deg(u) while the reflection weight is 2/ deg(u) - 1. Then on the tails, the dynamics is free because ... |
| [v8791] | ElliCE: Efficient and Provably Robust Algorithmic Recourse via the Rashomon Sets https://arxiv.org/abs/2602.07674 Robustness = 1 n n i=1 1 ∀f θ ∈ R(ε target ), f θ (x ci ) = c . A higher robustness score (closer to 1) is better, indicating that more counterfactual explanations are robust to model changes. Experimental Setup.For evaluators, we define a target m... |
| [v8861] | Distributed Network Application Security Policy Generation and Enforcement for Microsegmentation https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260067336).pn The method of claim 1, wherein the microsegmentation policy includes constraints applied during machine learning classification to optimize at least one of performance, accuracy, or human interpretability. 8. The method of claim 1, wherein the host ... |
| [v8965] | SYBR Green qPCR Master Mix manufacturer Echniques. https://www.siksinhibitor.com/2022/05/31/8570/ The authors in use state-of-the-art meta-learning schemes,namely MAML, FOMAML, REPTILE, and CAVIA, for IoT scenarios working with offline and on the internet meta finding out strategy. The outcomes show the benefit of meta-learning in both offline a... |
| [v8985] | The AI-native agency model is emerging across three major verticals of professional services. http://ai-native-agency.com/blog/ai-native-agency-verticals Sub-linear infrastructure scaling: Infrastructure costs (servers, API subscriptions, tooling) scale sub-linearly with revenue. Doubling the client base does not double infrastructure costs - it might increase them by 30-50%. The compounding effect o... |
| [v9083] | We describe an exact algorithm to solve linear systems of the form Hx = b where H is the Hessian of a deep net. https://doi.org/10.48550/arxiv.2601.06096 Unfortunately, there seems to exist no variant of Pearlmutter's trick to compute the Hessian-inverse-vector products directly. The proposed Hessian-inverse-vector product algorithm takes advantage of a deep net's layerwise structure.... |
| [v9141] | NutVLM: A Self-Adaptive Defense Framework against Full-Dimension Attacks for Vision Language Models in Autonomous Driving https://arxiv.org/abs/2602.13293 Furthermore, CADA utilizes risky scene induction to dismantle the causal reasoning required for navigation, encompassing both local and global adversarial threats. These evolving attacks underscore the urgent need for more effective defense methods.... |
| [v9145] | Opaque machine-learning models are systems whose internal decision logic is not directly interpretable by human stakeholders. https://www.ask.com/lifestyle/blackbox-ai-architectures-explainability-governance-considerations Robustness testing probes responses to distributional shift and adversarial perturbations. Fairness metrics check disparate impacts across groups. Explainability evaluation assesses fidelity (how well an explanation matches model behavior) and useful... |
| [v9146] | Versatile Behavior Diffusion for Generalized Traffic Agent Simulation https://doi.org/10.1109/tits.2026.3662886 Notably, our VBD model achieves this with fewer parameters than autoregressive generation models, achieving a balance between performance and computational efficiency. We present a selection of qualitative simulation results in Fig. 3, showcasing the... |
| [v9152] | Entropy-Regularized Token-Level Policy Optimization for Language Agent Reinforcement https://arxiv.org/abs/2402.06700 Besides, a reward signal is obtained after executing a complete action, which is too sparse to provide fine-grained supervision for each token.Applying it to all tokens within an action as Equation 5 might lead to a misalignment between token generat... |
| [v9156] | Publications by 'Chan Yeob Yeun' https://researchr.org/alias/chan-yeob-yeun Data Poisoning Against Federated Learning: Comparative Analysis Under Label-Flipping Attacks and GAN-Generated EEG DataMaryam Alsereidi, Abeer Awadallah, Alreem Alkaabi, Sangyoung Yoon, Chan Yeob Yeun. Investigating How Data Poising Attacks Can Impac... |
| [v9175] | In recommender systems, usually the ratings of a user to most items are missing and a critical problem is that the missing ratings are often missing not at random (MNAR) in reality. https://icml.cc/virtual/2019/session/4915 The ability to perform offline A/B-testing and off-policy learning using logged contextual bandit feedback is highly desirable in a broad range of applications, including recommender systems, search engines, ad placement, and personalized health care... |
| [v9237] | TAMAS: Benchmarking Adversarial Risks in Multi-Agent LLM Systems https://doi.org/10.48550/arXiv.2511.05269 An agent can invoke these tools to perform the user task. O = (o 1 , o 2 , . . . , o m ) denotes the observations based on the actions taken by the agents. For a given query q we aim to maximize: where a b is the benign action and 1 is an indicator f... |
| [v9344] | TeraSignal Introduces TSLink: Protocol-Agnostic Intelligent Interconnect for Plug-and-Play Linear Optics in AI Infrastructure https://www.prnewswire.com/news-releases/terasignal-introduces-tslink-protocol-agnostic-intelligent-interconnect-for-plug-and-play-linear-optics-in-ai-infrastructure-302250369.html Lower Bit Error Rate: TSLink eliminates the quantization noise introduced by analog-to-digital converters (ADCs) in DSP-based re-timers, significantly improving the BER in the link. Reduced Latency: TSLink removes the high latency caused by DSP proc... |
| [v9394] | Minimizing Hallucinations and Communication Costs: Adversarial Debate and Voting Mechanisms in LLM-Based Multi-Agents https://www.mdpi.com/2076-3417/15/7/3676 This paper aims to address the hallucination issue of LLMs by introducing adversarial and voting mechanisms in multi-agent LLMs.... |
| [v9402] | Blockchain Trends To Look Forward To in 2026 https://intellivon.com/blogs/blockchain-trends/ With continuous developments down the line, blockchain will act as the governance backbone for AI, logging every model version, dataset lineage, parameter change, and deployment approval on an immutable ledger. Smart contracts will enforce multi-part... |
| [v9482] | Most n8n AI agents fail in production. https://chronexa.io/blog/n8n-ai-agent-node-enterprise-architecture-guide-(2026) Crucially, production systems require confidence scoring and human-in-the-loop (HITL) thresholds. We implement logic that forces the agent to self-evaluate its output. If the extraction confidence falls below a pre-defined threshold - say 94% - the s... |
| [v9512] | OOWM: Structuring Embodied Reasoning and Planning via Object-Oriented Programmatic World Modeling https://arxiv.org/abs/2604.09580 First, it generates the State Abstraction ( state ), mapping visual features to a structured object hierarchy.Subsequently, it derives the Control Policy ( control ), which instantiates the Transition Logic (T ), governing the executable cleaning wor... |
| [v9514] | Chapter 10: Data Drift in LLMs - Causes, Challenges, and Strategies https://nexla.com/ai-infrastructure/data-drift/ Organizations must strategically plan their data collection efforts, seeking diverse sources and timely representation to bolster re-training initiatives. Data augmentation process (Source) #5 Dynamic adaptation Dynamic adaptation is continuous re... |
| [v9529] | In today's digital age, 5G technology has become the backbone of connectivity, supporting everything from mobile communications to smart cities and autonomous vehicles. https://moderndiplomacy.eu/2024/10/27/securing-5g-networks-how-ai-is-changing-the-game/ Integration with Security Information and Event Management (SIEM) tools allows for real-time threat detection and response, enhancing the network's resilience.... |
| [v9541] | Comparative Analysis of Statistical, Time - Frequency, and SVM Techniques for Change Detection in Nonlinear Biomedical Signals https://www.mdpi.com/2624-6120/5/4/41 By leveraging large-scale datasets and hierarchical representations, deep learning models can automatically learn discriminative features and detect subtle changes in signals with high accuracy. Moreover, techniques such as transfer learning and adve... |
| [v9614] | XiaoYee / Awesome_Efficient_LRM_Reasoning Public https://github.com/XiaoYee/Awesome_Efficient_LRM_Reasoning Meta-Reasoner: Dynamic Guidance for Optimized Inference-time Reasoning in Large Language Models Test-Time Preference Optimization: On-the-Fly Alignment via Iterative Textual Feedback TreeBoN: Enhancing Inference-Time Alignment with Speculative Tree... |
| [v9618] | Why do RAG systems fail at scale? https://www.kapa.ai/blog/rag-gone-wrong-the-7-most-common-mistakes-and-how-to-avoid-them What causes embedding rot and how do I fix it? Embedding rot occurs when the vector store remains static but the underlying data changes. Essentially, your responses will be based on stale data. Consider re-indexing your store when: 10-15% of your ... |
| [v9672] | MAPPO-LCR: Multi-Agent Proximal Policy Optimization with Local Cooperation Reward in spatial public goods games https://doi.org/10.1016/j.chaos.2026.117948 MAPPO is a Centralized-Training and Decentralized-Execution (CTDE) framework that extends the original PPO algorithm to cooperative multiagent systems. Let π θ (a i t | s i t ) denote the decentralized policy of agent i with parameters θ. Each agent ... |
| [v9689] | Explainable AI (XAI) refers to techniques and methods that make the behavior and outputs of artificial intelligence systems understandable to humans. https://www.respan.ai/glossary/explainable-ai The EU AI Act requires transparency for high-risk AI systems. GDPR's Article 22 gives individuals the right to meaningful information about automated decision-making logic. US regulations like ECOA and FCRA require explanations for adverse credit dec... |
| [v9717] | Home > Open Access Journals > MCA > Vol. 8 > Iss. https://digitalcommons.usf.edu/mca/vol8/iss1/8/ Blockchain technology in its most basic form is a distributed, immutable ledger that can be used to store data and is controlled by various nodes. By recording system activities and operational data on a distributed, tamper-evident blockchain, we dev... |
| [v9720] | Causal modeling of school aversion in psychiatrically referred adolescents: a DoWhy-based analysis https://pubmed.ncbi.nlm.nih.gov/41952142/ Causal inference was conducted through a combined framework of DAG learning, DoWhy estimation with backdoor propensity-score weighting and logistic-model-based counterfactual simulation. All analyses were performed using Python 3.11.8, with pgmpy, Do... |
| [v9728] | Think Locally, Explain Globally: Graph-Guided LLM Investigations via Local Reasoning and Belief Propagation https://arxiv.org/abs/2601.17915 LLM agents excel when environments are mostly static and the needed information fits in a model's context window, but they often fail in open-ended investigations where explanations must be constructed by iteratively mining evidence from massive, het... |
| [v9804] | Mira Network, a provider of decentralized AI infrastructure for trustless verified intelligence, has launched its testnet alongside a next generation suite of API's marking a major milestone in secur https://www.dlnews.com/research/internal/mira-network-launches-highly-anticipated-next-gen-suite-of-apis-and-testnet-for-verified-ai-intelligence/ Large language models (LLMs) and generative AI tools have revolutionized how people interact with technology, but they often grapple with challenges such as AI hallucinations and bias. Mira tackles these issues head-on with a novel distributed consen... |
| [v9929] | Toward Faithful Explanations in Acoustic Anomaly Detection https://doi.org/10.48550/arXiv.2601.12660 In this work, we study the interpretability of autoencoder-based models for audio anomaly detection, by comparing a standard autoencoder (AE) with a mask autoencoder (MAE) in terms of detection performance and interpretability. We applied several att... |
| [v9991] | Designing Human-Centered AI to Prevent Medication Dispensing Errors: Focus Group Study With Pharmacists https://pubmed.ncbi.nlm.nih.gov/38145475/ This study highlights the process of designing a human-centered AI for dispensing verification, emphasizing its interpretability, confidence visualization, and collaborative human-machine teaming styles. (2023)... |
| [v10050] | Safety Instincts: LLMs Learn to Trust Their Internal Compass for Self-Defense https://doi.org/10.48550/arXiv.2510.01088 We introduce Safety Instincts Reinforcement Learning (SIRL), which transforms this internal confidence into a self-generated reward signal, eliminating dependence on external validators or human annotations. SIRL teaches models to trust their safety ... |
| [v10165] | Soft actor-critic algorithm and improved GNN model in secure access control of disaggregated optical networks https://doi.org/10.1038/s41598-025-15225-z The study primarily tests the decision efficiency and communication overhead of GESAC under different network topology scales, assessing its scalability limit.The results are shown in Fig. 10: As shown in Fig. 10, the distributed architecture of GESA... |
| [v10170] | Interpretability refers to the degree to which human experts can understand and explain a system's decisions or outputs. https://www.xcubelabs.com/blog/explainability-and-interpretability-in-generative-ai-systems/ Feature attribution: Identifying which parts of the input image contributed to the generated output. Counterfactual explanations: Understanding how changes in the input image would affect the generated output. Model interpretability: Analyzing the ... |
| [v10273] | Modeling what Matters: Emergent Abstraction In Reinforcement Learning - Robotics Institute Carnegie Mellon University https://www.ri.cmu.edu/event/modeling-what-matters-emergent-abstraction-in-reinforcement-learning/ On the model-free, multi-agent side, we introduce Partial Reward Decoupling (PRD), a game-abstraction mechanism that dynamically decomposes teams into subgroups, simplifying cross-agent credit assignment and accelerating cooperative learning. We also... |
| [v10345] | Taming the Curses of Multiagency in Robust Markov Games with Large State Space through Linear Function Approximation https://arxiv.org/abs/2605.03125 Abstract: Multi-agent reinforcement learning (MARL) holds great potential but faces robustness challenges due to environmental uncertainty. To address this, distributionally robust Markov games (RMGs) optimize worst-case performance when the environm... |
| [v10351] | DPWriter: Reinforcement Learning with Diverse Planning Branching for Creative Writing https://huggingface.co/papers By leveraging diversity-seeking reinforcement learning algorithms, we introduce a novel sparse reward function for token-level learning signals that encourage diverse, high-likelihood latent CoT, overcoming deterministic sampling limitations and avoi... |
| [v10468] | typed-recall added to PyPI https://pypi.org/project/typed-recall/ Memory layer for AI agents - typed-edge graph, bounded hallucination, audit-grade, surgically forgettable. ... A B C, all supports edges True A B C with C A contradicts (frustrated triangle) False Pure-contradicts cycle False (frustration=1.00) ... |
| [v10524] | Introduce Chain-of-Model (CoM) paradigm to enhance scaling efficiency and inference flexibility. https://ainativefoundation.org/ai-native-daily-paper-digest-20250520/ Introduce AdaCoT (Adaptive Chain-of-Thought) to address inefficiencies in reasoning tasks for Large Language Models by adaptively determining when to invoke Chain-of-Thought. Utilize reinforcement learning with Proximal Policy Optimization to adjust... |
| [v10597] | How AI QA Teams Are Debugging the Future of Software Quality https://vmblog.com:443/archive/2025/07/16/how-ai-qa-teams-are-debugging-the-future-of-software-quality.aspx Software teams work with tight deadlines and complex systems. Manual testing can't always keep up - it happens late, misses edge cases, and doesn't scale well. ... ... severity and root cause Store data in centralized repositories accessible by you... |
| [v10619] | Highlights of all 1,899 NeurIPS-2020 papers. https://resources.paperdigest.org/2020/11/neurips-2020-highlights/ 99 Model-Based Multi-Agent RL In Zero-Sum Markov Games With Near-Optimal Sample Complexity Highlight: In this paper, we aim to address the fundamental open question about the sample complexity of model-based MARL. Related Papers Related Patents Rel... |
| [v10752] | Toward Safe and Human-Aligned Game Conversational Recommendation via Multi-Agent Decomposition https://doi.org/10.48550/arxiv.2504.20094 Finally, to mitigate safety and transparency risks (Challenge 3), MATCHA introduces a Risk Control Agent that detects adversarial prompts and filters harmful outputs, alongside an Explanation Agent that generates detailed, user-facing rationales to e... |
| [v10841] | Quantum Circuit Design for Training Perceptron Models https://arxiv.org/abs/1802.05428 In the appendix, we show that the success probability has a similar scaling with that of Gaussian distribution when the weight vector is unifromly sampled from the unit sphere of the version space, and it can be higher when the dimension of the versi... |
| [v10859] | Towards desiderata-driven design of visual counterfactual explainers https://doi.org/10.1016/j.patcog.2025.112811 Our in-the-loop gain evaluation can also be viewed as a simulation of a human study, with the difference that the user is modeled as an oracle and the study is fully reproducible.Furthermore, measuring performance gain rather than relying on subjecti... |
| [v10873] | CASC's Machine Intelligence Group was founded in 2020 to create a home base for technical staff and postdocs conducting fundamental and applied research in machine learning (ML) in support of the La https://computing.llnl.gov/casc/machine-intelligence-group Sam Sakla: deep learning, computer vision, self-supervised learning, fine-grained classification, object detection, manifold learning, multi-resolution image/signal processing Gautam Singh: generative models, large language models, agent learning, m... |
| [v10903] | Think Deep and Fast: Learning Neural Nonlinear Opinion Dynamics from Inverse Dynamic Games for Split-Second Interactions https://doi.org/10.1109/icra55743.2025.11127283 Outracing champion Gran Turismo drivers with deep reinforcement learning. P R Wurman, S Barrett, Nature. 6022022 Learn Thy Enemy: Online, Task-Aware Opponent Modeling in Autonomous Racing. L Chen, S Manuel, J Delgado, J Subotsis, P Tylkin, Symposium... |
| [v11003] | Language-Guided Multi-Agent Learning in Simulations: A Unified Framework and Evaluation https://doi.org/10.48550/arxiv.2506.04251 LLM-Communicator: Serves as a decentralized communication interface, enabling agents to encode, decode, and interpret emergent natural language messages for coordination.Agents exchange symbolic messages such as "cover me" or "focus fire" generated f... |
| [v11067] | PQS-BFL: A post-quantum secure blockchain-based federated learning framework https://doi.org/10.1016/j.eswa.2026.131449 This growth is sub-linear, suggesting that the system can handle an increasing number of clients without prohibitive increases in round duration, at least within the tested range.The average per-client transaction time remained relatively stable or e... |
| [v11082] | Cross-Modal Attention Analysis and Optimization in Vision-Language Models: A Study on Visual Reliability https://arxiv.org/abs/2604.17217 Future research directions include validating optimization strategies on natural image datasets, evaluating larger-scale VLMs, exploring explicit cross-modal alignment constraints such as contrastive loss regularization and attention guidance, develo... |
| [v11121] | Are You the A-hole? A Fair, Multi-Perspective Ethical Reasoning Framework https://arxiv.org/abs/2605.00270 We propose a neuro-symbolic aggregation framework that formalizes conflict resolution through Weighted Maximum Satisfiability (MaxSAT). Our pipeline utilizes a language model to map unstructured natural language explanations into interpretable logica... |
| [v11134] | Recent work in machine learning has yielded in algorithms with high performance and accuracy. https://projekter.aau.dk/performance-evaluation-of-explainable-ai-algorithms-against-adversarial-noise-03096450.html To overcome this issue, explainable AI (XAI) algorithms have been developed to add an extra layer of explainability towards AI. But with adversarial attacks at hand, even these algorithms become vulnerable. The aim of this paper is to study the effec... |
| [v11265] | Aligning Agent Policy with Externalities: Reward Design via Bilevel RL https://cdnjs.deepai.org/profile/mengdi-wang Parameter-Efficient Sparsity for Large Language Models Fine-Tuning With the dramatically increased number of parameters in language models,... 0 Yuchao Li, et al. ' Near-optimal Offline Reinforcement Learning with Linear Representation: Leveraging... |
| [v11311] | COHORT: Hybrid RL for Collaborative Large DNN Inference on Multi-Robot Systems Under Real-Time Constraints https://arxiv.org/abs/2603.10436 To move beyond single decision makers and enable collaborative execution across multiple edge devices, several works formulate task execution as a multi-agent control problem.In , edge servers are modeled as partially observable agents in a Dec-POMDP... |
| [v11321] | Learning Long-Context Diffusion Policies via Past-Token Prediction https://arxiv.org/abs/2505.09561 Recent research in language modeling, image generation, and robotics has shown that inference-time compute may allow models to improve their performance .Some seek to build an additional verifier to re-rank the output samples [9,17,41,42], while othe... |
| [v11337] | This paper introduces a novel XAI-based methodology to detect adversarial attacks on deepfake detectors. https://deepfake-demo.aisec.fraunhofer.de/related_work/2403.02955 The XAI-based approach effectively detects adversarial attacks on visual deepfake detectors, with Saliency and Guided Backpropagation generally yielding the highest accuracy, especially when the full model is finetuned. The method shows promising gen... |
| [v11347] | SpatiO: Adaptive Test-Time Orchestration of Vision-Language Agents for Spatial Reasoning https://arxiv.org/abs/2604.21190 SpatiO assembles a diverse pool of VLMs with distinct architectures, training objectives, and geometric inductive biases, each independently solving the spatial query under a designated reasoning role.We propose a novel Test-Time Orchestration (TTO) ... |
| [v11421] | In an era where identity is the new perimeter, we deploy cognitive security architectures that leverage real-time behavioral telemetry and autonomous policy enforcement to secure the enterprise at sc https://sabalynx.com/ai-identity-access-management/ The "Hard Truth" is managing the 8% margin of error. ""AI Hallucination" in IAM manifests as anomalous bypasses where the model misinterprets a legitimate but rare user behavior as a threat - or a sophisticated adversary's "low and slow" attack as be... |
| [v11683] | AI-Assisted Code Migration: 2026 Guide to Agentic Modernization https://article-realm.com/article/Computers/Software/82236-AI-Assisted-Code-Migration-2026-Guide-to-Agentic-Modernization.html The smartest enterprises we've seen build human-in-the-loop (HITL) checkpoints at every critical decision point - especially for business logic transformations, security-sensitive code, and regulatory compliance sections. Our investigation demonstra... |
| [v11707] | Artificial Intelligence Selection And Configuration https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260127494).pn Artificial Intelligence Selection And Configuration --- The method of claim 5, wherein the AI component type optimized for data storage or retrieval comprises a blockchain-based distributed ledger, wherein automatically configuring the intelligent ag... |
| [v11756] | Online Topology Inference from Streaming Stationary Graph Signals with Partial Connectivity Information https://doi.org/10.3390/a13090228 Indeed, we examine how the variability and eigenvectors of the underlying graph as well as the diffusion filters' frequency response influence the size of the convergence radius (or misadjustment in the adaptive filtering parlance). (2020)... |
| [v11766] | Submitted on 27 May 2019 (v1), last revised 4 Oct 2019 (this version, v2)] https://arxiv.org/abs/1905.11468v2 First, we derive new per-image theoretical robustness bounds based on local gradient information. These bounds strongly motivate input gradient regularization. Second, we implement a scaleable version of input gradient regularization which avoids dou... |
| [v11794] | Towards Assessing and Benchmarking Risk-Return Tradeoff of Off-Policy Evaluation Haruka https://speakerdeck.com/harukakiyohara_/towards-risk-return-assessment-of-ope May 2024 Towards assessing risk-return tradeoff of OPE 12 (estimated) marginal importance weight state-action visitation probability Summary of OPE Off-Policy Evaluation (OPE) aims to evaluate the expected performance of a policy using only offline... |
| [v11819] | PointMAC: Meta-Learned Adaptation for Robust Test-Time Point Cloud Completion https://doi.org/10.48550/arxiv.2510.10365 A meta-auxiliary learning strategy based on Model-Agnostic Meta-Learning (MAML) ensures that adaptation driven by auxiliary objectives is consistently aligned with the primary completion task.During inference, we adapt the shared encoder on-the-fly b... |
| [v11850] | Persistent cognitive machine with curated long term memory https://patents.google.com/?oq=19321173 These adapters handle variations in formatting, vocabulary, and reasoning granularity, ensuring smooth thought transfer between models with different characteristics. The cache incorporates a contextual validation layer that assesses thought applica... |
| [v11937] | In this article: View the comprehensive list of regulations available to build assessments in Compliance Manager. https://learn.microsoft.com/en-us/purview/compliance-manager-regulations-list ISO/IEC 23894:2023 ISO/IEC 42001:2023 NIST AI Risk Management Framework (RMF) 1.0 Guidelines and Functional Requirements for Electronic Records Management Systems (ICA Module 2) ISO 15489-1:2016 ISO 16175-1:2020 ISO 19791 - Information technolo... |
| [v11938] | Temporal Action Proposal Generation with Background Constraint - NewsBreak https://www.newsbreak.com/news/2462358269144/temporal-action-proposal-generation-with-background-constraint ... for Self-Supervised Visual Pre-Training - https://newsbreak.com/news/2463395356139/masked-feature-prediction-for-self-supervised-visual-pre-training URL: Constraints on subleading interactions in beta decay Lagrangian - https://newsbreak.com/new... |
| [v11946] | Generation-Augmented Latent Navigation for Continuous Spatiotemporal Zoom and Rotation in Immersive Environments https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260017457).pn Generation-Augmented Latent Navigation for Continuous Spatiotemporal Zoom and Rotation in Immersive Environments --- The system further incorporates a symbolic anchor manager that establishes persistent semantic landmarks within the latent space, ena... |
| [v11995] | We've observed that in applied RL settings, the question of whether it makes sense to use multi-agent algorithms often comes up. https://rise.cs.berkeley.edu/blog/scaling-multi-agent-rl-with-rllib/ Similarly, policy-gradient algorithms like A3C and PPO may struggle in multi-agent settings, as the credit assignment problem becomes increasingly harder with more agents. Consider a traffic gridlock between many autonomous agents. It is easy to see ... |
| [v12013] | Multi-Agent Systems and Optimization: Enhancing Efficiency Through Collaborative AI https://smythos.com/developers/agent-development/multi-agent-systems-and-optimization/ By leveraging advanced algorithms and distributed decision-making, MAS have demonstrated their ability to outperform traditional approaches in areas such as traffic management and energy distribution. The power of MAS lies in their ability to break ... |
| [v12056] | The effect of data poisoning on counterfactual explanations https://doi.org/10.1016/j.inffus.2026.104237 This work studies the vulnerability of counterfactual explanations to data poisoning.We formalize data poisoning in the context of counterfactual explanations for increasing the cost of recourse on three different levels: locally for a single instanc... |
| [v12070] | D-REX: A Benchmark for Detecting Deceptive Reasoning in Large Language Models https://doi.org/10.48550/arXiv.2509.17938 We define this as a scenario where a model produces a benign or helpful response, while its internal reasoning process, or chain-of-thought (CoT), follows a hidden, malicious directive. This behavior can be induced by sophisticated system prompt inje... |
| [v12098] | Neural Rendering For Inverse Graphics Generation https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260127820).pn In at least one embodiment, and without limitation, machine learning models used by system may include machine learning model(s) using linear regression, logistic regression, decision trees, support vector machines (SVM), Naive Bayes, k-nearest neigh... |
| [v12118] | Getting value from your data shouldn’t be this hard https://www.technologyreview.com/2021/10/19/1037290/getting-value-from-your-data-shouldnt-be-this-hard/ As data's applications grow and become more ubiquitous, producers, consumers, and owners and stewards of data are finding that they don't have a playbook to follow. Consumers want to connect to data they trust so they can make the best possible decis... |
| [v12122] | AegisMCP: Online Graph Intrusion Detection for Tool-Augmented LLMs on Edge Devices https://doi.org/10.48550/arXiv.2510.19462 Robust training (edge-dropout, adversarial negatives), conservative novelty weighting, and guardrail escalators for high-risk motifs (e.g., install then egress to a new domain) reduce susceptibility. Topology-aware regularization and adversarial subg... |
| [v12125] | Federated Learning (FL) is a distributed learning paradigm that leverages the computational strength of local devices to collaboratively train a model. https://scholarsmine.mst.edu/comsci_facwork/2048/ The clients train the local model on their respective devices and submit the weight updates to the server for aggregation. This paradigm allows the clients to experience diverse data without sharing their local data with other participants or the ser... |
| [v12128] | Interplay between Security, Privacy and Trust in 6G-enabled Intelligent Transportation Systems AHMED DANLADI ABDULLAHI * (Student Member, IEEE), ERFAN BAHRAMI † , TOOSKA DARGAHI * (Member, IEEE), https://doi.org/10.48550/arxiv.2510.02487 Dynamic trust computation in multi-agent systems Computing and adapting trust scores for vehicles in dynamic, adversarial, and high-mobility settings remains underexplored, particularly for large-scale, real-world ITS deployments. significant privac... |
| [v12130] | Machine Learning (ML) continues to evolve rapidly, driven by advances in hardware, model architectures, and data-centric methodologies. https://dev.to/ashishsinghbora/a-technical-deep-dive-into-machine-learning-architectures-paradigms-and-optimization-strategies-cpd Automated retraining via CI/CD pipelines, feature stores (e.g., Feast), and model registries (e.g., MLflow, SageMaker). Hybrid deployment models combining serverless inference, on-prem acceleration, and edge serving. Neuro-Symbolic and Hybrid AI C... |
| [v12143] | e-Postgraduate Diploma (ePGD) in Computer Science And Engineering https://www.mygreatlearning.com/iit-bombay-e-postgraduate-diploma-computer-science-engineering The course then develops expertise in value-based methods, including their extension using function approximation and deep learning for complex, high-dimensional environments. It further covers different classes of RL methods such as policy-gradient ... |
| [v12162] | ARES: Adaptive Red-Teaming and End-to-End Repair of Policy-Reward System https://arxiv.org/abs/2604.18789 ... blind spots and biases. The second stage then utilizes this improved RM to optimize the Core LLM, creating a more robustly aligned system overall. Extensive experiments across diverse safety evaluations demonstrate that ARES substantially improve... |
| [v12165] | CiteAudit: You Cited It, But Did You Read It? A Benchmark for Verifying Scientific References in the LLM Era https://arxiv.org/abs/2602.23452 We design a multi-agent verification pipeline that decomposes citation checking into metadata extraction, memory lookup, web-based retrieval, and final judgment. To evaluate this, we construct a large-scale, human-validated dataset spanning diverse d... |
| [v12184] | fairadapt: Causal Reasoning for Fair Data Pre-processing https://arxiv.org/abs/2110.10200 The following sections describe an implementation of the fair data adaptation method outlined in Plecko and Meinshausen (2020), which combines the notions of counterfactual fairness and resolving variables, and explicitly computes counterfactual valu... |
| [v12212] | FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning https://arxiv.org/abs/2511.14715 The reliability threshold Θ t at round t evolves based on model convergence and detected anomalies: where Θ base is the baseline threshold, conv(w t ) measures model convergence (higher values indicate stable training), and anomaly rate t represents ... |
| [v12225] | Blockchain-based federated learning methodologies in smart environments https://doi.org/10.1007/s10586-021-03424-y Blockchain-based federated learning methodologies in smart environments --- In , authors combined Blockchain technology and FL using Python, creating Biscotti with the goal of privacy and maintaining the accuracy of FL at the same time. In FL, there ... |
| [v12247] | Preserving Causal Constraints in Counterfactual Explanations for Machine Learning Classifiers https://arxiv.org/abs/1912.03277 A key question for the Oracle-based method is the number of labelled CF examples it needs. Using the Adult dataset and the non-decreasing Age constraint, we show the Constraint-Feasibility Score of OracleGenCF as we increase the number of labelled CF... |
| [v12260] | Therefore, a well-defined and robust knowledge base (correctly structuring the syntax and semantic rules of the respective domain) is vital in allowing the machine to generate logical conclusions th http://www.eectod.com/%E0%B8%82%E0%B9%88%E0%B8%B2%E0%B8%A7%E0%B8%9B%E0%B8%A3%E0%B8%B0%E0%B8%8A%E0%B8%B2%E0%B8%AA%E0%B8%B1%E0%B8%A1%E0%B8%9E%E0%B8%B1%E0%B8%99%E0%B8%98%E0%B9%8C/the-third-wave-of-artificial-intelligence-neuro/ How to explain the input-output behavior, or even inner activation states, of deep learning networks is a highly important line of investigation, as the black-box character of existing systems hides system biases and generally fails to provide a rati... |
| [v12261] | The AI Agent Stability Gap: Why Your AI Agents Fail in Production (2026) https://hyperion-consulting.io/de/insights/ai-research-decoded-the-2026-stability-gap-what-s-holding-back-your-ai-agents GDPR compliance: Supports on-device fine-tuning (via LoRA), allowing adaptation to specific voices/faces without external data sharing. Data requirement: Training demands 1,000+ hours of labeled audio-video data per domain. Public datasets (e.g., Vo... |
| [v12267] | Adversarial machine learning https://en.wikipedia.org/?curid=45049676 An attacker may poison this data by injecting malicious samples during operation that subsequently disrupt retraining. Data poisoning techniques can also be applied to text-to-image model s to alter their output, which is used by artists to defend th... |
| [v12284] | This course book is protected by copyright. https://studylib.net/doc/26236460/blockchain ... record keeping, consensus, independent validation, and an immutable ledger. not all distributed ledgers are implemented with blockchain, blockchain is the primary... |
| [v12298] | EvoCurr: Self-evolving Curriculum with Behavior Code Generation for Complex Decision-making https://arxiv.org/abs/2508.09586 EvoCurr: Self-evolving Curriculum with Behavior Code Generation for Complex Decision-making --- with their corresponding types and abilities, environmental settings including map and terrain features, task objectives that define win conditions and ev... |
| [v12311] | Thanks to Advait Jayant (Peri Labs), Sven Wellmann (Polychain Capital), Chao (Metropolis DAO), Jiahao (Flock), Alexander Long(Pluralis Research), Ben Fielding & Jeff Amico (Gensyn), for their insigh https://0xjacobzhao.substack.com/p/the-holy-grail-of-crypto-ai-frontier Gensyn's RL Swarm enables decentralized coordination in the post-training phase. Each node runs its own model locally - no gradient synchronization required - allowing efficient operation in heterogeneous, unstable environments. Its workflow mimics R... |
| [v12340] | AI-Powered Optimization of Supply Chain Operations https://www.ibtimes.co.in/ai-powered-optimization-supply-chain-operations-883640 Effective solutions build strong data pipelines and assign specialized teams to eliminate silos. Equally vital is computational efficiency - especially in time-sensitive functions. Hybrid cloud-edge architectures have addressed latency and reliabilit... |
| [v12355] | A Survey of Slow Thinking-based Reasoning LLMs using Reinforced Learning and Inference-time Scaling Law https://arxiv.org/abs/2505.02665 Xie et al. proposed Guided Beam Search that conducts self-assessment at each step of the beam search algorithm to guide the selection of promising reasoning paths. REINFORCED LEARNING In this section, we summarize the related studies of reinforced... |
| [v12392] | NuGet\Install-Package QuantumSuperposition -Version 1.9.0 https://www.nuget.org/packages/QuantumSuperposition Generic superposition engine for QuBit and Eigenstates: arithmetic, comparisons and LINQ style queries over many possible values at once with complex weights, sampling, entanglement and non observational operations. Physics flavoured quantum system:... |
| [v12403] | Graph Defense Diffusion Model https://doi.org/10.1145/3770854.3780207 Graph Neural Networks (GNNs) are highly vulnerable to adversarial attacks, which can greatly degrade their performance. Existing graph purification methods attempt to address this issue by filtering attacked graphs.... |
| [v12421] | An earlier version of this post is on the RISELab blog. https://bair.berkeley.edu/blog/2018/12/12/rllib/ Similarly, policy-gradient algorithms like A3C and PPO may struggle in multi-agent settings, as the credit assignment problem becomes increasingly harder with more agents.... |
| [v12449] | JudgeMeNot: Personalizing Large Language Models to Emulate Judicial Reasoning in Hebrew https://arxiv.org/abs/2604.18041 In contrast, doubling the rank yields only a modest +0.77 BLEU increase and negligible changes in semantic and style scores. These results indicate diminishing returns from increasing adapter rank, while additional training examples continue to impro... |
| [v12472] | Resilient Multi-Dimensional Consensus and Distributed Optimization against Agent-Based and Denial-of-Service Attacks https://arxiv.org/abs/2510.06835 On the one hand, adversarial agents including malicious, Byzantine, or stubborn ones can drive the normal agents' states outside the desired region . On the other hand, attacks launched at the communication links, such as DoS attacks, can prevent inf... |
| [v12525] | A Unified Framework for Evaluating and Enhancing the Transparency of Explainable AI Methods via Perturbation-Gradient Consensus Attribution https://arxiv.org/abs/2412.03884 Second, we introduce Perturbation-Gradient Consensus Attribution (PGCA), which fuses grid-based perturbation importance with Grad-CAM++ through consensus amplification and adaptive contrast enhancement, combining perturbation fidelity with gradient-b... |
| [v12549] | A dual-layered robust design optimization framework for nonlinear assembly processes using uncertainty-aware deep ensemble and metaheuristic algorithms https://doi.org/10.2139/ssrn.6255261 By integrating Deep Ensemble with Monte Carlo Dropout, the proposed model not only provides precise multi-target predictions for six performance metrics but also quantifies aleatoric and epistemic uncertainties, ensuring high predictive reliability i... |
| [v12560] | GitHub - erwanlemerrer/awesome-audit-algorithms: A curated list of algorithms and papers for auditing black-box algorithms. https://github.com/erwanlemerrer/awesome-audit-algorithms Auditing fairness under unawareness through counterfactual reasoning - (Information Processing & Management) Shows how to unveil whether a black-box model, complying with the regulations, is still biased or not. XAudit : A Theoretical Look at Auditi... |
| [v12585] | Adaptive Collaboration of Arena-Based Argumentative LLMs for Explainable and Contestable Legal Reasoning https://arxiv.org/abs/2602.18916 Crucially, our framework supports a Human-in-the-Loop (HITL) contestability workflow, enabling users to directly audit and modify the underlying reasoning graph to influence the final judgment. Empirical evaluations on the LegalBench benchmark demons... |
| [v12624] | Weakest Link in the Chain: Security Vulnerabilities in Advanced Reasoning Models https://arxiv.org/abs/2506.13726 However, this overall trend masks significant category-specific differences: for certain attack types the reasoning models are substantially more vulnerable (e.g., up to 32 percentage points worse on a tree-of-attacks prompt), while for others they a... |
| [v12699] | Resilient Dynamic Average Consensus based on Trusted agents https://doi.org/10.48550/arxiv.2303.08171 Next we define a connectivity property of the graph from . Definition 1 (Connected Dominating Set (CDS)): A set S of graph Γ = (V, E) is a CDS if all nodes belonging to S form a connected graph, and each node which does not belong to S has at least ... |
| [v12723] | Tree-of-Reasoning: Towards Complex Medical Diagnosis via Multi-Agent Reasoning with Evidence Tree https://doi.org/10.48550/arXiv.2508.03038 Considering that most of the existing medical datasets are singlesource medical data, To evaluate different methods under complex medical diagnosis scenario, we collect real patient data from a realworld hospital, which included patient information (... |
| [v12791] | Center for Information and Language Processing https://doi.org/10.48550/arxiv.2305.14250 Additionally, it performs joint reasoning across answer candidates and operates at a much larger scale (e.g., over 350 nodes on average for each question) and with a variety of constraint types. REFLEX: Our Approach Belief Graphs Our belief graphs... |
| [v12800] | Privacy-Preserving Federated Learning with Adaptive Noise Scaling and Enhanced CNN Models https://doi.org/10.37745/ejcsit.2013/vol13n52126137 Differential privacy (DP) provides formal guarantees but often degrades performance, especially in non-independent and identically distributed (non-IID) settings. This work proposes an adaptive noise scaling mechanism to integrate DP into FL more eff... |
| [v12837] | Adaptive homomorphic federated learning framework for multi-institutional medical imaging with optimized diagnostic accuracy https://pubmed.ncbi.nlm.nih.gov/42082627/ NASFL combines multi-level homomorphic encryption (MLHE) and stochastic differential privacy to provide patient confidentiality while using a transformer-guided ResNet backbone for adaptive multi-modal feature fusion between X-ray and CT imaging data... |
| [v12842] | The meeting will be held virtual through Microsoft Teams. https://slim.gatech.edu/content/ML4Seismic-Partners-Meeting-Fall-2021 Bayesian inference for ill-posed inverse problems is challenged by the high-dimensionality of the unknown, computationally expensive forward operator, and choosing a prior distribution that accurately encodes prior knowledge on the unknown. To handle... |
| [v12851] | glacier-creative-git/knowledge-graph-traversal-semantic-rag-research: Completed research on semantic retrieval augmented generation through novel knowledge graph traversal algorithms https://github.com/glacier-creative-git/similarity-graph-traversal-semantic-rag-research ... for all metrics. This is due to its agnosticism towards the original query; it only traverses based on relevancy to the current chunk. This explains the significant underperformance in 20qa-themes-gpt4omini-reasoning, particularly in faithfulness... |
| [v12874] | Self-Aware Vector Embeddings for Retrieval-Augmented Generation: A Neuroscience-Inspired Framework for Temporal, Confidence-Weighted, and Relational Knowledge https://arxiv.org/abs/2604.20598 Feedback poisoning: an adversary who can submit positive feedback can inflate confidence; rate-limits, feedback-source weighting, and anomaly detection are needed. Ripple runaway: dense graphs risk cascade explosion; the hard D max bound and per-hop ... |
| [v12898] | Multi-Timescale, Gradient Descent, Temporal Difference Learning with Linear Options https://arxiv.org/abs/1703.06471 Deliberating on large or continuous state spaces have been long standing challenges in reinforcement learning. Temporal Abstraction have somewhat made this possible, but efficiently planing using temporal abstraction still remains an issue. Moreover ... |
| [v12899] | Data science: a natural ecosystem https://doi.org/10.1016/j.inffus.2025.104113 Data science: a natural ecosystem --- For this, certain theoretical assumptions on the underlying model are needed.Predictive modeling has been widely adopted by the empirical machine learning community.Donoho argues that the secret sauce boosting p... |
| [v12910] | Human-AI Use Patterns for Decision-Making in Disaster Scenarios: A Systematic Review https://doi.org/10.1109/istas65609.2025.11269624 By improving transparency in the AI decision-making process, their study demonstrated that human operators could better understand system behavior, which reduced over-reliance and led to more accurate and contextually grounded decisions.This reinforc... |
| [v12930] | Towards desiderata-driven design of visual counterfactual explainers https://doi.org/10.1016/j.patcog.2025.112811 Visual counterfactual explainers (VCEs) are a straightforward and promising approach to enhancing the transparency of image classifiers. ... Similar to methods such as DiffeoCF , ACE , and DiME , we ensure a focus on plausible data transformation x →... |
| [v12954] | On the Convergence of Single-Timescale Actor-Critic https://doi.org/10.48550/arxiv.2410.08868 Our analysis shows a sample complexity of O(ϵ -3 ) to compute an ϵ-optimal policy, improving upon the prior best rate of O(ϵ -4 ). ODE-Based Methodology with Direct Global Guarantees: Our core technical innovation is a streamlined ODE-based analysi... |
| [v12976] | Sub-optimality bounds for certainty equivalent policies in partially observed systems https://arxiv.org/abs/2602.02814 For models where the cost and the dynamics are smooth in an appropriate sense, we derive upper bounds on the sub-optimality of certainty equivalent policies.We present several examples to illustrate the results. I. INTRODUCTION In many applications... |
| [v12977] | Protein Counterfactuals via Diffusion-Guided Latent Optimization https://arxiv.org/abs/2603.10811 Translating counterfactual methods to proteins introduces two fundamental challenges.First, the manifold constraint: Unlike images, proteins are governed by strict epistatic constraints -a single core mutation can abolish folding while a compensatory... |
| [v12981] | Towards Fine-Grained Interpretability: Counterfactual Explanations for Misclassification with Saliency Partition https://doi.org/10.1109/cvpr52734.2025.02797 To address this limitation, we propose a fine-grained counterfactual explanation framework that generates both objectlevel and part-level interpretability, addressing two fundamental questions: (1) which fine-grained features contribute to model misc... |
| [v12993] | bartCause is an R package that uses Bayesian Additive Regression Trees (BART) to adjust for confounding variables without making parametric assumptions. https://thinkcausal.org/en/page/bart-cause/ If we can appropriately model the outcome, we can impute missing counterfactual outcomes and then find our causal estimates. thinkCausal uses BART for causal inference, taking advantage of its non-parametric, flexible approach to outcome modeling. W... |
| [v13005] | Robust Explainability: A tutorial on gradient-based attribution methods for deep neural networks https://doi.org/10.1109/MSP.2022.3142719 Robust Explainability: A tutorial on gradient-based attribution methods for deep neural networks --- In the literature, the terms, attribution, relevance, importance, contribution, sensitivity, and saliency scores are synonymously used. Perturbation-... |
| [v13015] | Tech Mahindra announced collaboration with Microsoft to launch an ontology-driven Agentic AI platform that accelerates telecom and enterprise data modernization. https://digitalterminal.in/tech-companies/tech-mahindra-collaborates-with-microsoft-to-launch-ontology-driven-agentic-ai-platform Tech Mahindra announced collaboration with Microsoft to launch an ontology-driven Agentic AI platform that accelerates telecom and enterprise data modernization. 07 Mar 2026, 5:42 am Built on Microsoft Fabric and Azure AI Foundry, the solution enab... |
| [v13037] | Artificial Intelligence will be used to accelerate new medicine discovery in a University of Liverpool partnership secured following Mayor Steve Rotheram's US trade mission. https://news.liverpool.ac.uk/2026/02/05/new-university-of-liverpool-us-collaboration-to-accelerate-drug-discovery-using-ai/ Our collaboration with BPGbio, Inc. brings together cutting-edge Bayesian computation, multi-omics research, and secure data environments to deliver exactly that. This is the blueprint for the next generation of precision medicine." Niven R. Narain,... |
| [v13048] | Unifying Adversarial Perturbation for Graph Neural Networks https://doi.org/10.48550/arXiv.2509.00387 Specifically, these methods mainly apply perturbation to the node feature, weights or graph structure. suggest dropping edges randomly in adversarial training to generate perturbations on the adjacency matrix A. designs a dynamic regularizer forcin... |
| [v13053] | Non-Intrusive Load Monitoring Model Based on SimCLR and Visualized Color V-I Trajectories https://pubmed.ncbi.nlm.nih.gov/41755171/ Initially, unlabeled load data from the source domain (PLAID) and target domain (WHITED) are converted into RGB color V-I trajectories and input into the model. The framework enhances intra-class aggregation through contrastive learning and achieves... |
| [v13054] | Tokenization of Intellectual Property (IP) https://reddit.com/r/BuildOnWYZth/comments/1hv1v1s/tokenization_of_intellectual_property_ip/ Enhance transparency and trust through blockchain's immutable ledger. * Enable broader access to IP investment opportunities.... |
| [v13128] | Dual-Modal Lung Cancer AI: Interpretable Radiology and Microscopy with Clinical Risk Integration https://arxiv.org/abs/2604.16104 Explainable AI techniques including Grad-CAM, Grad-CAM++, Integrated Gradients, Occlusion, Saliency Maps, and SmoothGrad are applied to provide visual interpretability.... |
| [v13129] | Towards East Asian Facial Expression Recognition in the Real World: A New Database and Deep Recognition Baseline https://www.mdpi.com/1424-8220/22/21/8089 Deep learning methods such as convolutional neural networks (CNN) , deep belief networks (DBN) ,deep autoencoders (DAE) , and generative adversarial networks (GAN) are gradually gaining popularity among researchers. CNN relies on a set of learnable ... |
| [v13135] | Reinforcement Learning for Decision-Level Interception Prioritization in Drone Swarm Defense https://doi.org/10.48550/arxiv.2508.00641 The rapid proliferation of unmanned aerial vehicles has spurred a surge in research on autonomous defense systems capable of detecting, prioritizing, and neutralizing aerial threats, particularly in swarm-based attack scenarios.These efforts span mul... |
| [v13163] | In an era where data privacy concerns increasingly shape public acceptance of digital health technologies, a new study states that advanced AI does not have to come at the cost of patient confidentia https://www.devdiscourse.com/article/technology/3791526-privacy-first-ai-models-bring-breakthrough-in-iot-based-healthcare Errors tend to occur in borderline cases, such as early-stage disease or intermediate biomarker values, highlighting the importance of integrating AI outputs with clinical decision support rather than using them in isolation. This reinforces the view... |
| [v13176] | GoDaddy Inc.: DEF 14A (DEF 14A) https://www.sec.gov/Archives/edgar/data/0001609711/0001609711-26-000030-index.htm 2025 Peer Group Akamai Technologies, Inc. (NASDAQ: AKAM) Autodesk, Inc. (NASDAQ: ADSK) Docusign, Inc. (NASDAQ: DOCU) eBay Inc. (NASDAQ: EBAY) Fortinet, Inc. (NASDAQ: FTNT) Gen Digital Inc. (NASDAQ: GEN) HubSpot, Inc. (NYSE: HUBS) Nutanix, Inc. (NASDA... |
| [v13179] | Toward Individual Fairness Without Centralized Data: Selective Counterfactual Consistency for Vertical Federated Learning https://arxiv.org/abs/2605.07117 Our focus is on individual-level counterfactual stability, i.e., per-instance prediction consistency under protected-attribute interventions as formalized in the causal fairness literature, rather than group parity guarantees such as demographic pari... |
| [v13206] | SkillGraph: Self-Evolving Multi-Agent Collaboration with Multimodal Graph Topology https://arxiv.org/abs/2604.17503 Conditioning the topology predictor on textual agent profiles alone is therefore insufficient. To capture this visual dependency, we introduce the Multimodal Graph Transformer (MMGT), a five-stage encoder that jointly processes image patches, questio... |
| [v13219] | Employ Blockchain to Boost Cloud Computing Cybersecurity: Product Data Integrity and Appropriate Access with Smart Contract Regulations https://doi.org/10.1109/ICTBIG68706.2025.11323968 With blockchain-based decentralized, append-only, immutable ledger and smart contract programmability, the architecture supports secure data sharing, auditable trails, enforceable access rule automation that is not dependent on central parties. The b... |
| [v13235] | Article: Virtual Panel: What to Consider when Adopting Large Language Models https://www.infoq.com/articles/llm-adoption-considerations/ For a lot of enterprises, their LLM applications will be touching fairly business-sensitive data, and for them it may be important that they control the model that sees that data. Secondly, customizability. When you self-host models you control all ... |
| [v13262] | Constructive Distortion: Improving MLLMs with Attention-Guided Image Warping https://doi.org/10.48550/arXiv.2510.09741 Finally, note that we intervene before feature extraction, while the above methods operate after the image has already been encoded, often from features that have already lost critical spatial detail (Pantazopoulos et al., 2024). In summary, our key ... |
| [v13265] | Efficient Low-Rank GNN Defense Against Structural Attacks https://doi.org/10.1109/ickg59574.2023.00006 Many approaches to defend GNNs against adversarial attacks have been proposed.Some works utilize pre-processing methods to filter the perturbed graph structure prior to the training stage , . (2023)... |
| [v13275] | Building Trustworthy AI by Addressing its 16+2 Desiderata with Goal-Directed Commonsense Reasoning https://doi.org/10.48550/arxiv.2506.12667 2 Background: s(CASP) s(CASP), by Arias et al. (2018), is a novel non-monotonic reasoner that evaluates Constraint Answer Set Programs without a grounding phase either before or during execution.s(CASP) supports predicates and thus retains logical va... |
| [v13307] | From Load Tests to Live Streams: Graph Embedding-Based Anomaly Detection in Microservice Architectures https://arxiv.org/abs/2604.06448 Does introducing a synthetic load along a selected call path improve anomaly detection evaluation?Answering this required careful design, as injecting synthetic anomalies is inherently nontrivial.Naively adding noise can yield ambiguous results, espe... |
| [v13333] | I recently released "Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting" with collaborators Julian Michael, Ethan Perez, and Sam Bowman. https://www.lesswrong.com/posts/6eKL9wDqeiELbKPDj/unfaithful-explanations-in-chain-of-thought-prompting I recently released "Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting" with collaborators Julian Michael, Ethan Perez, and Sam Bowman. In this post, I briefly elaborate on motivations/implication... |
| [v13336] | Deep Reinforcement Learning for Decentralized Multi-Robot Exploration With Macro Actions https://doi.org/10.1109/lra.2022.3224667 Between MDPs and semi-MDPs: A framework for temporal abstraction in reinforcement learning. R S Sutton, D Precup, S Singh, Artif. Intell. 1121/2R. S. Sutton, D. Precup, and S. Singh, "Between MDPs and semi-MDPs: A framework for temporal abstraction i... |
| [v13375] | Circular Economy and Green Environment https://www.mdpi.com/journal/ijerph/special_issues/Circular_Economy_Green_Environment To obtain a thorough understanding and explanation of the influencing mechanism of environmental regulation (ER) on green innovation efficiency (GIE), the super-slack based measure-data envelopment analysis (Super-SBM-DEA) method was applied to evalu... |
| [v13405] | CDC Workshop on Decentralization in Teams and Games, Dec 2025. https://adityam.github.io/talks.html CDC Workshop on Decentralization in Teams and Games, Dec 2025. Agent-state based policies in POMDPs: Beyond belief-state MDPs (slides) (video) ... Sub-optimality bounds for certainty equivalence policies in POMDPs (slides) CDC Workshop on Decentral... |
| [v13407] | Machine learning-based discovery of informative SNPs for population assignment through whole genome sequencing https://doi.org/10.1186/s12864-025-12322-1 M E Hossain, M A Kabir, L Zheng, D L Swain, S Mcgrath, J Medway, Artif Intell Agric. 62022 Classification and regression by randomforest. A Liaw, M Wiener, Forest. 232001 Support Vector Machines * the interface to libsvm in package e1071. D Meyer, ... |
| [v13414] | Adversarial Robustness in AI-Driven Cybersecurity Solutions: Thwarting Evasion Assaults in Real-Time Detection Systems https://doi.org/10.22161/ijaems.115.9 Malicious entities create subtle alterations in network traffic or system actions that mislead AI models into misidentifying threats as harmless, facilitating evasion tactics that can circumvent real-time intrusion detection systems (IDS). This study... |
| [v13444] | Discover how social media verification methods inspire robust AI authenticity practices to build trust and model integrity. https://fuzzypoint.net/how-to-verify-authenticity-in-ai-systems-insights-from-media Yes, which is why cryptographic anchoring and continuous adversarial testing are crucial for maintaining model integrity. How does user trust improve with AI transparency? When AI systems explain their processes clearly and allow user feedback, tru... |
| [v13478] | Real-Time Distributed Model Predictive Control with Limited Communication Data Rates. (arXiv:2208.12531v2 [eess.SY] UPDATED) http://arxiv.org/abs/2208.12531 ... multi-agent systems (MASs) necessitates communication between agents, yet the consequence of communication data rates is typically overlooked. This work focuses on developing stability-guaranteed control methods for MASs with limited data rate... |
| [v13496] | The phenomenon of multimodal LLM hallucination represents one of the most critical challenges facing the deployment of large vision-language models in real-world applications. https://www.libertify.com/interactive-library/multimodal-llm-hallucination-survey/ A model might describe objects not present in an image, assign wrong colors or sizes to visible objects, or fabricate spatial relationships that contradict the actual visual scene. These hallucinations pose substantial obstacles to practical deployme... |
| [v13727] | Human-computer interaction (HCI) is a multidisciplinary field of study that focuses on how people interact with technology. https://computing.njit.edu/human-computer-interaction-0 Research Areas: human-AI teaming, interactive visualization, visual analytics, responsible AI, humanmachine communication Human-AI Collaboration using Visual Analytics... |
| [v13729] | The Hessian of tall-skinny networks is easy to invert https://doi.org/10.48550/arXiv.2601.06096 Given a way to compute the Hessian-vector product, one can indirectly compute the Hessian-inverse-vector product via, say Krylov iterations like Conjugate Gradient as proposed by Pearlmutter and more recently re-investigated . However, the quality of... |
| [v13741] | System And Method For Improved Structural Discovery And Representation Learning Of Multi-agent Data https://worldwide.espacenet.com/patent/search?q=EP4034962B1 The present disclosure generally relates to a system, non-transitory computer readable medium, and method for learning player distribution and role assignments in sports. Background Increasingly, sports fans and data analysts have become entrenched... |
| [v13743] | Learning to Defend by Attacking (and Vice-Versa): Transfer of Learning in Cybersecurity Games https://doi.org/10.1109/eurospw59978.2023.00056 The result is a model inspired by both bounded rationality and ToM. Experimental results comparing this model with a strategy that attempts to optimally learn to maximize utility, the upper confidence bound model, demonstrates the benefit of the prop... |
| [v13807] | Bipedal Action Model For Humanoid Robot https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260124750).pn These systems lack the temporal consistency needed for smooth, long-horizon tasks and are not robust enough to adapt to the unpredictable nature of real-world environments.... |
| [v13839] | by Jan Betley, Owain_Evans https://www.lesswrong.com/posts/ifechgnJRtJdduFGC/emergent-misalignment-narrow-finetuning-can-produce-broadly I'd be interested in knowing more about how the fine-tuning is regularized and the strength of any KL-divergence-penalty-ish terms. I'm not clear on how the openai fine-tuning API works here with default hypers. By default, I would expect that optim... |
| [v13867] | Ev-Trust: A Strategy Equilibrium Trust Mechanism for Evolutionary Games in LLM-Based Multi-Agent Services https://doi.org/10.48550/arXiv.2512.16167 Unlike traditional static or centralized reputation systems, Ev-Trust redefines trust as a dynamic and self-organizing process that drives strategic adaptation in open multi-agent ecosystems. By embedding both direct and indirect trust into agents' e... |
| [v13875] | Towards Explainable Federated Learning: Understanding the Impact of Differential Privacy https://doi.org/10.48550/arxiv.2602.10100 For instance, a malicious FL server can run a Gradient Inversion or a Membership Attack to obtain sensitive data. In order to achieve both, data privacy and explainability, this paper proposes a FL solution, called Federated EXplainable Trees with... |
| [v13878] | Abstract (296) HTML (9) PDF (2950KB)(1687) Knowledge map Save https://www.joca.cn/EN/article/showDownloadTopList.do Then, by establishing the SGAM (Spatial Global relationship Attention Module) and CGAM (Channel Global Attention Module), the spatial global relationship mechanism and channel attention mechanism were introduced to capture global information, so as t... |
| [v13909] | "domain": "Prompt Injection & Jailbreak Defense", "concept": "Probabilistic Output Manipulation via Logit Probing", "difficulty": "Hard", "text": "Explain how an attacker can perform a 'Jailbreak by https://huggingface.co/datasets/Roman1111111/gemini-3.1-pro-hard-high-reasoning ### DEFENSE ARCHITECTURE: Recursive Epistemic Gating (REG) **Concept:** Treat the Chain-of-Thought (CoT) not as a continuous generation stream, but as a series of atomic, verifiable transactions. The model is effectively "paused" after every newline ... |
| [v13930] | Hybrid Agentic AI and Multi-Agent Systems in Smart Manufacturing https://doi.org/10.1016/j.jmsy.2026.04.002 In contrast, Small Language Models (SLMs) offer a lightweight, privacy-preserving complement.Deployed locally on edge devices or factory nodes, SLMs can provide lowlatency reasoning, rapid diagnostics, and continuous monitoring without reliance on ex... |
| [v13947] | AI is about to put a whole new spin on virtual communication https://www.inverse.com/innovation/how-smart-replies-could-improve-socially-distanced-communications AI-mediated communication (AI-MC) represents a new paradigm where communication is augmented or generated by an intelligent system. As AI-MC becomes more prevalent, it is important to understand the effects that it has on human interactions and inter... |
| [v13976] | Trust-Based Assured Sensor Fusion in Distributed Aerial Autonomy https://doi.org/10.48550/arxiv.2507.17875 Thus, UAV data fusion needs specialized trust frameworks-to the best of our knowledge, none existed before this work. Trust-Based Fusion with Bayesian Principles We formulate a joint problem of trust estimation and sensor fusion using a hidden Mark... |
| [v14059] | 12.6.2025 Paper discussion: InstaSHAP: Interpretable Additive Models Explain Shapley Values Instantly. http://tml.cs.uni-tuebingen.de/teaching/tml_graduate_seminar/past_tml_graduate_seminar.php 9.2.2022 (paper discussion) Denoising Diffusion Probabilistic Models pdf, helpful blog posts here and here, Jonathan Ho, Ajay Jain, Pieter Abbeel, 2020.... |
| [v14084] | PatientEase - Domain-Aware RAG for Rehabilitation Instruction Simplification https://doi.org/10.3390/bioengineering12111204 A summary table that follows lays out each stripped version next to the full model for easy comparison Table 3.An ablation experiment confirms that the PatientEase system's inner components perform unique, non-replaceable roles.The user-situated retr... |
| [v14162] | Enabling verifiability in federated learning utilizing zero-knowledge proofs and blockchain https://doi.org/10.1109/AIAHPC66801.2025.11290017 To address the absence of process-level verifiability in federated learning, a verifiable architecture, zero-knowledge proof-verified and blockchain-audited federated learning (zk-BcFed), is proposed by integrating zero-knowledge proofs with blockcha... |
| [v14177] | MedRule-KG: A Knowledge-Graph-Steered Scaffold for Reliable Mathematical and Biomedical Reasoning https://doi.org/10.48550/arXiv.2511.12963 The monotonic increase in EM with dataset size further indicates that improvements are not artifacts of small-sample variability. Moreover, the flattening of the curve for the KG + Verifier system suggests saturation at high performance, implying tha... |
| [v14183] | Imagine you are a loan officer faced with a model that says "deny" for a borrower's application. https://legacy.thenextgentechinsider.com/flex-unlocking-feature-importance-with-counterfactual-explanations/ Computational cost Counterfactual generation ≈ O(N C) + cheap aggregation; comparable to sampling-based SHAP for modest C Sampling-based SHAP ≈ O(N S) with S ≈ 100-200 model queries Very cheap locally (one linear fit), but must be repeated for many n... |
| [v14190] | Comorbidity Classification from Clinical Free-Text using Large Language Models: Application to Sleep Disorder Patients https://doi.org/10.1007/s10916-026-02343-y The evaluation presented in this study is computational in nature and was conducted on prospectively scored comorbidity annotations.As a first study of its kind within this dataset, it is intended to lay the methodological foundation and provide init... |
| [v14201] | Provable Defense Framework for LLM Jailbreaks via Noise-Augumented Alignment https://arxiv.org/abs/2602.01587 This approach preserves the positional indices of the retained tokens and maintains the structural integrity of the prompt without introducing foreign tokens into the vocabulary.We present theoretical guarantee in Appendix. Noise-Augmented Alignment... |
| [v14244] | TRAM: Bridging Trust Regions and Sharpness Aware Minimization https://arxiv.org/abs/2310.03646 We propose Trust Region Aware Minimization (TRAM), a SAM algorithm fine-tuning for low parameter sharpness and smooth, informative representations preserving pre-trained structure. TRAM uses a trust region bound to inform the SAM adversarial neighbor... |
| [v14295] | DVD: Dynamic Contrastive Decoding for Knowledge Amplification in Multi-Document Question Answering https://doi.org/10.18653/v1/2024.emnlp-main.266 Prior research in RAG has introduced various improvements (Vu et al., 2023), such as improving retrieval quality (Shi et al., 2023d;Xu et al., 2023), refining responses through multiple iterations (Peng et al., 2023;Li et al., 2024), using optimized ... |
| [v14358] | Lost in Decoding? Reproducing and Stress-Testing the Look-Ahead Prior in Generative Retrieval https://doi.org/10.1145/3805712.3808567 Planning Ahead in Generative Retrieval (PAG) mitigates this failure mode by using simultaneous decoding to compute a document-level look-ahead prior that guides subsequent sequential decoding. We reproduce PAG at inference time and stress-test its de... |
| [v14366] | The Architectural Evolution of Intelligence: A Formal Taxonomy of the AI Technology Stack https://www.c-sharpcorner.com/article/the-architectural-evolution-of-intelligence-a-formal-taxonomy-of-the-ai-technol/ A* Search applies an admissible heuristic function h(n) one that never overestimates the true cost to guide best-first expansion of a state-space graph, guaranteeing optimal path discovery in O(b^d) time complexity where b is the branching factor and... |
| [v14404] | We generate a data set with 5,000 observations assigned over 5 equally sized batches, with 10 covariates and 4 treatment arms. https://ftp2.osuosl.org/pub/cran/web/packages/banditsCI/vignettes/banditsCI.html ... main = paste0("Assignment for arm ", k)) graphics::abline(v=cumsum(batch_sizes_w), col="#00ccff") graphics::legend("topleft", legend = 1:K, col=1:K, lty=1:K, lwd = 3) Estimating response. We then generate augmented inverse probability weighte... |
| [v14411] | Challenges in Credit Assignment for Multi-Agent Reinforcement Learning in Open Agent Systems https://doi.org/10.48550/arXiv.2510.27659 For the empirical analyses, we evaluate two representative algorithms, i.e., Deep Q-Network (DQN) for TCA, and Multi-Agent PPO (MAPPO) for SCA, respectively. Each method is adapted to operate in an environment with openness. To measure the impact o... |
| [v14441] | The Overfocusing Bias of Convolutional Neural Networks: A Saliency-Guided Regularization Approach https://arxiv.org/abs/2409.17370 Our SGDrop framework leverages attribution methods to regularize neural network training by selectively dropping the most salient pieces of information.Crucially, it is designed to be universally applicable and remains agnostic to the specific choice... |
| [v14442] | MARVEL: A Multi Agent-based Research Validator and Enabler using Large Language Models https://doi.org/10.48550/arxiv.2601.03436 It scores on a 0-1 scale for relevance and factual correctness relative to both the question and the provided context, with higher scores awarded for responses that cite evidence and a score of 0 assigned to responses that state an inability to answe... |
| [v14482] | Spatial Lifting for Dense Prediction https://doi.org/10.48550/arxiv.2507.10222 Providing reliable estimates of prediction uncertainty or quality is vital for deploying models in critical applications.Common approaches include Monte Carlo dropout , forming ensembles of models, or developing explicitly Bayesian neural networks, a... |
| [v14581] | Foundation Models for Causal Inference via Prior-Data Fitted Networks https://arxiv.org/abs/2506.10914 Then, we propose a concrete instantiation using Bayesian neural networks and provide a learning algorithm that leverages the SCM's ability to simulate counterfactual data and perform consistent Bayesian inference in a wide range of causal inference s... |
| [v14584] | LLM Inference Enhanced by External Knowledge: A Survey https://doi.org/10.48550/arXiv.2505.24377 These hybrid methods leverage the strengths of both symbolic and neural reasoning to overcome the limitations of either approach, making them particularly suitable for complex reasoning. Knowledge Graph (KG) Integration KG integration approaches var... |
| [v14668] | F Common Vulnerabilities in Internet of Things Security and How to Address Them? - https://www.thenetworkdna.com/2025/07/common-vulnerabilities-in-internet-of.html A concise, detailed answer explains that the discipline blends traditional network controls with device-specific safeguards such as signed bootloaders, low-power encryption ciphers, and life-cycle-aware asset tracking. Anchoring your strategy to that... |
| [v14694] | FORT-IDS: a federated, optimized, robust and trustworthy intrusion detection system for IIoT security https://doi.org/10.1038/s41598-025-31025-x The federated experiments in this paper therefore report round-wise behaviour under a many-client non-IID setting with K = 20 clients and client fraction C = 0.2 and show FedAvg aggregated accuracy converging to 0.934 by round five under our leakage-... |
| [v14739] | Large Language Models Encode Semantics and Alignment in Linearly Separable Representations https://arxiv.org/abs/2507.09709 1), though compression patterns vary by architecture and do not universally follow the U-shaped trends reported in prior work (Ansuini et al., 2019;Valeriani et al., 2023;Razzhigaev et al., 2024;Skean et al., 2025). Geometric encoding of alignment: i... |
| [v14855] | Mediation analysis to identify causes of racial disparity in health outcomes: a comparison of model-based and outcome-based approaches https://doi.org/10.1186/s12874-026-02776-6 The estimator for PA is:5 The standard error of the PA is estimated using the Delta method, a general method for deriving the variance of a function of asymptotically normal random variables with known variance. This estimation incorporates counterfa... |
| [v14893] | FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning https://arxiv.org/abs/2511.14715 FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning --- FLARE integrates: (i) a multi-dimensional reputation score capturing performance consistency, statistical anomaly indicators, and temporal behavior, ... |
| [v14894] | Dell Technologies is on the lookout for an AI-ML Engineer MCP-Agentic to fill the vacancy in its Hyderabad office. https://www.analyticsinsight.net/job-openings/ai-ml-engineer-mcp-agentic-dell Apply multi-agent orchestration to allow for self-governing decision-making and task assigning. Train AI models for identifying attacks, spotting deviations, and conducting user behavioral study. Establish guidelines for AI observability, monitorin... |
| [v14955] | Toward a Graph-Theoretic Model of Belief: Confidence, Credibility, and Structural Coherence https://doi.org/10.48550/arXiv.2508.03465 In this framework, each node represents an individual belief, while edges encode epistemic relationships-such as support, contradiction, or qualification-between beliefs. Crucially, each belief is endowed with two distinct attributes: credibility, wh... |
| [v15041] | The silent infrastructure: How Hassan's AI systems are quietly redefining cloud defense https://www.digitaljournal.com/tech-science/the-silent-infrastructure-how-hassans-ai-systems-are-quietly-redefining-cloud-defense/article Transparent audit flags to ensure human interpretability of alerts Security systems should not become surveillance systems, Hassan writes.... |
| [v15053] | Amplification of formal method and fuzz testing to enable scalable assurance for communication system https://patents.google.com/?oq=18628625 The method of claim 1, further comprising a step of establishing dependency relationships through cross-attention mechanisms and/or self-attention mechanisms. ... The amplification of the formal method and fuzz testing provides a general approach to ... |
| [v15059] | Integrating Reinforcement Learning with Visual Generative Models: Foundations and Advances https://doi.org/10.48550/arXiv.2508.10316 Key contributions include MADDPG , which introduced centralized training with decentralized execution, allowing agents to condition their critics on global information during training while executing independently at test time. Other approaches, such... |
| [v15123] | AI Triage Failure: When Moving Fast Becomes a Risk | HackerNoon https://hackernoon.com/ai-triage-failure-when-moving-fast-becomes-a-risk The Shift : From AI Projects to AI Products After those failures, we hit reset. We stopped thinking of AI as a "proof of concept" or "quick win." We started treating it like any long-living product - with versions, feedback loops, governance, and a... |
| [v15126] | A Roadmap towards Intelligent Operations for Reliable Cloud Computing Systems https://doi.org/10.48550/arxiv.2310.00677 Although cloud management frameworks provide automatic mechanisms for failure recovery, unplanned service failures may still cause severe cascading effects.Therefore, it is crucial to evaluate the impact of service failures rapidly and accurately for... |
| [v15154] | Tri-LLM Cooperative Federated Zero-Shot Intrusion Detection with Semantic Disagreement and Trust-Aware Aggregation https://doi.org/10.48550/arXiv.2602.00219 In contrast to centralized systems that frequently degrade under heterogeneous data distributions, the proposed Tri-LLM framework maintains consistent performance even when client semantics vary substantially. This robustness arises from semantic ali... |
| [v15167] | Primary focus: planning and shipping a production - ready chatbot integration powered by LLMs (e.g., OpenAI API) that becomes a real business asset - not a lab demo. https://towerhousestudio.com/blog/ai-chatbot-implementation-strategy/ List assumptions and dependencies that could delay delivery. Define acceptance criteria and exit criteria for the pilot. Data and retrieval. Which sources will be indexed and how access is granted. How sensitive data is handled, chunked, embedded, f... |
| [v15179] | MIRROR: A Multi-Agent Framework with Iterative Adaptive Revision and Hierarchical Retrieval for Optimization Modeling in Operations Research https://doi.org/10.48550/arXiv.2602.03318 Systems like Chain-of-Experts (Xiao et al., 2023), OptiMUS (Ahmaditeshnizi et al., 2024), and ORMind (Wang et al., 2025) decompose complex modeling tasks into specialized roles and enable iterative interaction among agents, offering a flexible and pr... |
| [v15224] | Finding and fixing a harmful behavior that WAS represented in the SAE training data in a way that is competitive with appropriate fine-tuning and machine unlearning baselines. https://www.lesswrong.com/posts/HYkg6kwqhCQT5uYuK/eis-xv-a-new-proof-of-concept-for-useful-interpretability Finding and fixing a harmful behavior that WAS CONVINCINGLY NOT represented in the SAE training data in a way that is competitive with appropriate fine-tuning and machine unlearning baselines. The reward model sycophancy behavior was developed by th... |
| [v15305] | The Dual Role of Abstracting over the Irrelevant in Symbolic Explanations: Cognitive Effort vs. Understanding https://arxiv.org/abs/2602.03467 Just as image classification explanations use saliency maps to highlight relevant pixels while treating the rest as irrelevant (Ribeiro et al., 2016), symbolic representations must distinguish between essential logical pivots and distracting details ... |
| [v15313] | TranSimHub:A Unified Air-Ground Simulation Platform for Multi-Modal Perception and Decision-Making https://doi.org/10.48550/arXiv.2510.15365 Dynamic entities include vehicles, pedestrians, and UAVs, which are controlled through predefined engines such as SUMO, or alternatively by user-defined strategies. Both ground and aerial agents support policy-level customization, allowing integratio... |
| [v15343] | In my previous blog, we explored the evolution of information retrieval techniques from simple keyword matching to sophisticated context understanding and introduced the concept that sparse embedding https://dev.to/zilliz/exploring-bge-m3-and-splade-two-machine-learning-models-for-generating-sparse-embeddings-22p1 "Learned" sparse embeddings are an advanced type of embedding that combines the precision of traditional sparse embeddings with the semantic richness of dense embeddings. They enhance the sparse retrieval approach by incorporating contextual informat... |
| [v15368] | "Learnings from Paying Artists Royalties for AI-Generated Art: A Retrospective on Tess.Design, Our Attempt to Make an Ethical, Artist-Friendly AI Marketplace. https://gwern.net/doc/ai/nn/diffusion/index "Learnings from Paying Artists Royalties for AI-Generated Art: A Retrospective on Tess.Design, Our Attempt to Make an Ethical, Artist-Friendly AI Marketplace. ... DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 S... |
| [v15436] | scGCN is a graph convolutional networks algorithm for knowledge transfer in single cell omics - News Break https://www.newsbreak.com/news/2288228997400/scgcn-is-a-graph-convolutional-networks-algorithm-for-knowledge-transfer-in-single-cell-omics In this work, we use these graph measures to explore the robustness of various ANNs to adversarial attacks. To this end, we (1) explore the design space of inter-layer and intra-layers connectivity regimes of ANNs in the graph domain and record their... |
| [v15437] | AgentRx: Diagnosing AI Agent Failures from Execution Trajectories https://doi.org/10.48550/arXiv.2602.02475 The list of recorded failures gives a causal chain from the first unrecoverable failure to the terminal one. A Cross-Domain Failure Taxonomy Prior work takes a system-level view of multi-agent failures, organizing failure modes by design, coordinati... |
| [v15455] | Moscow Exchange to Follow up BTC Futures Launch With Crypto Funds, Structured Bonds | MEXC News https://www.mexc.com/lv-LV/news/21251 In the entire AI Agent protocol stack, we divided it into three main layers in our previous research report, namely Agent Infrastructure Layer: This layer provides the lowest-level operational support for agents and is the technical foundation for al... |
| [v15471] | Method And System For Recording And Enforcing Encumbrances On Assets Using Multiple Secure, Immutable Ledgers https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260127563).pn FIG. depicts an exemplary distributed ledger similar to the hybrid distributed ledger environment as shown in FIG. . The example distributed ledger includes a public distributed ledger layer including a blockchain having blocks - of transactions. In ... |
| [v15478] | We introduce 2D-Malafide, a novel and lightweight adversarial attack designed to deceive face deepfake detection systems. https://www.eurecom.fr/fr/publication/7876 We introduce 2D-Malafide, a novel and lightweight adversarial attack designed to deceive face deepfake detection systems. ... Additionally, we report an explainability analysis using GradCAM which illustrates how 2D-Malafide misleads detection syste... |
| [v15586] | Light management for image and data control https://patents.google.com/?oq=17555507 Light management for image and data control --- This is implementer optional and adjustable and is analogous to the graduating effect of a bright spot removal process wherein "darkening" corrections (LRC actions) that are more peripheral to the centr... |
| [v15822] | Agent health score for agentic automations https://patents.google.com/?oq=19216203 For instance, AI agents make use of generative AI models. Generative AI models can generate various types of content, such as text, imagery, audio, and synthetic data. Various types of generative AI models may be used, including, but not limited to, ... |
| [v15831] | Reactive Multi-agent Coordination using Auction-based Task Allocation and Behavior Trees https://doi.org/10.1109/ccta54093.2023.10252961 Behavior trees also generalize other popular control structures, such as finite state machines and decision trees , thus increasing its utility as a flexible and versatile framework for automation. C. Contributions With respect to the aforementione... |
| [v15838] | 4 Oct 202566B23F41159AB61353DF219B4E3FE4ADarXiv:2510.03612v1[cs.AI]User query: "Find a Thriller Movie" https://doi.org/10.48550/arxiv.2510.03612 Recent studies reveal that these agents are vulnerable against attackers who can bias selection outcomes through preference manipulations using adversarial pop-ups, image perturbations, or content tweaks.Existing work, however, either assumes strong ... |
| [v15909] | Quantum-Inspired Neural Network with Sequence Input () https://scirp.org/journal/paperinformation Ref. proposed a neural network model with quantum gated nodes and a smart algorithm for it, which shows superior performance in comparison with a standard error back propagation network. Ref. proposed a weightless model based on quantum circuit. It... |
| [v15921] | This week in deep learning, we bring you Tensorflow Similarity, faster quantized inference with XNNPACK, the world's first 5G and AI enabled drone platform and a paper on transformer-based 3D dance g https://www.deeplearningweekly.com/p/deep-learning-weekly-issue-215 A comprehensive introduction to Optimum, an optimization toolkit that provides performance optimization tools targeting efficient AI hardware and built-in collaboration with hardware partners. CARLA: A Python Library to Benchmark Algorithmic Recours... |
| [v16000] | LLM/Agent-as-Data-Analyst: A Survey https://doi.org/10.48550/arxiv.2509.23988 The Extractor-Reasoner-Executor paradigm extracts relevant context, generates logic rules or equations, and executes them via LLM prompting to get the final answer.Similarly, S3HQA uses a retriever to filter heterogeneous resources, a selector to i... |
| [v16027] | SocialJax: An Evaluation Suite for Multi-agent Reinforcement Learning in Sequential Social Dilemmas https://doi.org/10.48550/arxiv.2503.14576 However, using a common reward structure can exacerbate the credit assignment problem.Specifically, if an agent takes an arbitrary action concurrently with a teammate who performs a successful action generating a reward, the agent may mistakenly attr... |
| [v16044] | DocSync: Agentic Documentation Maintenance via Critic-Guided Reflexion https://arxiv.org/abs/2605.02163 DocSync bridges syntactic changes and natural language descriptions by fusing Abstract Syntax Tree (AST) representations and Retrieval-Augmented Generation (RAG) to provide dependency-aware context. Furthermore, to ensure factual consistency, we inco... |
| [v16046] | Throughout this essay, I use "mathematical fluency" to mean something specific: not manual derivations or rote memorization, but structural literacy - the ability to recognize when seemingly disparat https://www.insights.phyusionbio.com/p/the-end-of-disciplinary-sovereignty Techniques originally developed in one field are rapidly generalized and redeployed elsewhere. Causal discovery methods from econometrics now inform drug target identification. Transformer architectures - initially designed for natural language proce... |
| [v16089] | Generative Image Layer Decomposition with Visual Effects https://doi.org/10.1109/cvpr52734.2025.00716 Petru-Daniel Tudosiu, Yongxin Yang, Shifeng Zhang, Fei Chen, Steven Mcdonagh, Gerasimos Lampouras, Ignacio Iacobacci, Sarah Parisot, Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. the IEEE/CVF Conference on Compute... |
| [v16090] | A Comprehensible Explanation of the Dimensions in CNNs - News Break https://www.newsbreak.com/news/2289464574587/a-comprehensible-explanation-of-the-dimensions-in-cnns In this paper, we introduce a novel framework that harnesses explainable ML methods to guide high-fidelity assessment of ML evasion attacks. Our framework enables explanation-guided correlation analysis between pre-evasion perturbations and post-evas... |
| [v16104] | 12,000+ API Keys and Passwords Found in Public Datasets Used for LLM Training https://thehackernews.com/2025/02/12000-api-keys-and-passwords-found-in.html Such adversarial attacks are called prompt injections, which occur when an attacker manipulates a generative artificial intelligence (GenAI) system through crafted inputs, causing the LLM to unknowingly produce otherwise prohibited content. Recent f... |
| [v16149] | This package shows how to multiply the inverse of the Hessian of a deep network with a vector. https://vuink.com/post/tvguho-d-dpbz/a-rahimi/hessian Pearlmutter showed a clever way to compute the Hessian-vector-product for a deep net. By contrast, the paper and code in this repo shows how to compute the Hessian-inverse-product, the product of the inverse of the Hessian of a deep net with a vector... |
| [v16190] | Individual Contributions as Intrinsic Exploration Scaffolds for Multi-agent Reinforcement Learning https://doi.org/10.48550/arxiv.2405.18110 ... z t ), known as the noisy TV problem (Schmidhuber, 2010).Our focus is primarily on the individual contribution r i t,int , which necessitates a specific measurement method to effectively distinguish the contribution of agent i's action u i t and ... |
| [v16195] | Detecting Adversarial Data via Perturbation Forgery https://doi.org/10.48550/arXiv.2405.16226 Although previous detection methods achieve high performance in detecting gradient-based adversarial attacks, new attacks based on generative models with imbalanced and anisotropic noise patterns evade detection. Even worse, existing techniques eithe... |
| [v16222] | Amplification of formal method and fuzz testing to enable scalable assurance for communication system https://patents.google.com/?oq=18628625 ... have been identified in these networks. To perform safety-critical tasks at scale, swarms of autonomous aerial drones should be capable of rapidly reconfiguring and adapting in degraded conditions and reliably detecting and recovering from advers... |
| [v16242] | Probabilistic Perspectives on Error Minimization in Adversarial Reinforcement Learning https://doi.org/10.48550/arXiv.2406.04724 Deep Reinforcement Learning (DRL) policies are highly susceptible to adversarial noise in observations, which poses significant risks in safety-critical scenarios. For instance, a self-driving car could experience catastrophic consequences if its sen... |
| [v16245] | AI-Based System and Method for Generating Enhanced Radiology Reports https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260128138).pn According to one embodiment, the report integration module is configured to integrate the AI-generated radiology report into a patient's electronic health record (EHR) using standards such as Health Level Seven (HL7), Fast Healthcare Interoperability... |
| [v16289] | Abstract: This article surveys the current state of artificial intelligence - what it can and cannot do today - across theory, technologies, representative applications, limitations, and governance. https://www.upuply.com/blog/what-can-ai-do-today For generative media, the trade-off between fidelity and controllability matters: higher fidelity generative models can create convincing audio and video, but controlling specifics (e.g., consistent character motion across scenes) remains difficult, ... |
| [v16323] | Adversarial Examples (AI)Adversarial TrainingAI EvaluationsDeceptive AlignmentMachine Learning (ML)AI https://www.lesswrong.com/posts/oPnFzfZtaoWrqTP4H/solving-adversarial-attacks-in-computer-vision-as-a-baby Despite my fundamental belief that machines can (eventually) do anything, the human brain seems to have some particularly great solutions to many challenging problems, especially where robustness extending to very rarified, long tails is needed (such... |
| [v16338] | Edge-Intelligent Block Chain Framework for Federated Privacy-Preserving Medical Diagnostics https://doi.org/10.1109/IC2NC67409.2025.11376420 The framework also employs an energy-optimized consensus mechanism using adaptive Practical Byzantine Fault Tolerance (PBFT) to improve transaction throughput and scalability in edge environments. Experimental evaluation using the MIMIC-III and Physi... |
| [v16376] | FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning https://doi.org/10.48550/arXiv.2511.14715 The server performs the entire multi-dimensional reputation assessment Section III-B and dynamic thresholding III-C on these noisy updates.... |
| [v16401] | Dynamic Allostery of the Catabolite Activator Protein Revealed by Interatomic Forces https://pubmed.ncbi.nlm.nih.gov/26244893/ For full activation and DNA binding, the homodimeric protein requires the binding of two cyclic AMP (cAMP) molecules in an anti-cooperative manner, the source of which appears to be largely of entropic nature according to previous experimental studie... |
| [v16416] | Universal Soldier: Using Universal Adversarial Perturbations for Detecting Backdoor Attacks https://doi.org/10.1109/DSN-W60302.2024.00024 This is similar to universal adversarial perturbations (UAP). Indeed, UAPs are input-agnostic perturbations capable of misleading a well-trained model. We observe an intuitive phenomenon: UAPs generated from backdoored models need fewer perturbations... |
| [v16438] | Decision Transparency Enhancement And Integration Of User Feedback And Control Of Artificial Intelligence Outputs https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260127199).pn Decision Transparency Enhancement And Integration Of User Feedback And Control Of Artificial Intelligence Outputs --- The system of claim 1, wherein the natural language response comprises at least one explanation type selected from the group consist... |
| [v16446] | Prophet, Revisited: Practical Time-Series Forecasting at Scale https://joshuaberkowitz.us/blog/github-repos-8/prophet-revisited-practical-time-series-forecasting-at-scale-847 Design choices emphasize interpretability and guardrails. Trend changepoints are regularized to prevent overfitting; seasonalities are represented with Fourier series; and holidays enter as binary regressors. The Python API mirrors scikit-learn's fi... |
| [v16468] | Exploration in Deep Reinforcement Learning: From Single-Agent to Multiagent Domain https://doi.org/10.1109/tnnls.2023.3236361 The high entropy of TV becomes an irresistible attraction to the agent. In Fig. 4, we show a similar 'Noisy-TV' in VizDoom on the right. The uncontrollable Gaussian noise is added to the observation space, which attracts the agent to stay in the cur... |
| [v16482] | FASE : A Fairness-Aware Spatiotemporal Event Graph Framework for Predictive Policing https://arxiv.org/abs/2604.18644 The absence of baselines means we cannot claim predictive superiority over simpler approaches. Fairness metric limitations.The DIR constraint measures patrol-intensity parity, not outcome parity.As demonstrated in Section 4.3, allocation-level DIR ≈... |
| [v16509] | Most multi-agent AI systems fail at coordination, not capability. https://particula.tech/blog/multi-agent-ai-orchestration-that-works The single biggest source of multi-agent system failures is unstructured communication. When agents pass free-form text to each other, small phrasing changes cause downstream misinterpretations that cascade through the system. Define Typed Message S... |
| [v16526] | Galaxy vs UFO ² vs Linux Agent vs Mobile Agent: When to Use What? https://microsoft.github.io/UFO/project_directory_structure/ Event-Driven Coordination Safe Assignment Locking Agent Output Observer Using as Galaxy Device Speculative Multi-Action Windows Agent Arena Markdown Log Viewer Windows App Environment Creating Custom MCP Servers Creating Custom Third-Party A... |
| [v16531] | A Quantum-Resistant and AI-Resilient Real-Time Keystroke Protection Framework With Blockchain-Backed Decentralized Identity https://doi.org/10.1109/ACCESS.2026.3680275 The system integrates Hyperledger Fabric for tamper-evident mapping management, W3C Decentralized Identifier (DID) support for self-sovereign identity, and optional zero-knowledge authentication to eliminate password transmission. Session keys are de... |
| [v16556] | Are Foundation Models All You Need for Zero-shot Face Presentation Attack Detection? http://www.visionbib.com/bibliography/update/2601.html ... computationally efficient framework leveraging auxiliary head features for robust cloth-changing person re-identification, A Concentration Inequalities for Semidefinite Least Squares Based on Data Concept-Based Explanation for Deep Vision Model... |
| [v16569] | Bayesian Active Inference for Intelligent UAV Anti-Jamming and Adaptive Trajectory Planning https://doi.org/10.48550/arXiv.2512.05711 This paper proposes a hierarchical trajectory planning framework for UAVs operating under adversarial jamming conditions. Leveraging Bayesian Active Inference, the approach combines expert-generated demonstrations with probabilistic generative modeli... |
| [v16615] | The Role of Blockchain in Zero Trust Architecture | HackerNoon https://hackernoon.com/the-role-of-blockchain-in-zero-trust-architecture Third, a blockchain-based log of network events offers a tamper-evident audit trail, elevating the concept of " verify everything " to an unassailable record of transactions and actions. Given that Zero Trust involves continuous monitoring, having an... |
| [v16647] | Prototype Learning for Explainable Brain Age Prediction https://doi.org/10.1109/WACV57701.2024.00772 Explainable Brain Age Prediction: Several studies have attempted to introduce explainability into brain age prediction models, predominantly for adult MRI. Saliency methods have been used to explain brain age predictions [9,21,28,30,50], but their ex... |
| [v16658] | Trust-Aware AI-Enabled Edge Framework for Intelligent Traffic Control in Cyber-Physical Systems https://www.techscience.com/results Abstract The rapid evolution of smart cities has led to the deployment of Cyber-Physical IoT Systems (CPS-IoT) for real-time monitoring, intelligent decision-making, and efficient resource management, particularly in intelligent transportation and ve... |
| [v16662] | Dynamic Adversarial Fine-Tuning Reorganizes Refusal Geometry https://arxiv.org/abs/2604.27019 Abstract: Safety-aligned language models must refuse harmful requests without collapsing into broad over-refusal, but the training-time mechanisms behind this tradeoff remain unclear. Prior work characterizes refusal directions and jailbreak robustne... |
| [v16678] | Zero-Shot Policy Transfer in Multi-Agent Reinforcement Learning via Trusted Federated Explainability https://doi.org/10.63282/3050-9246.ijetcsit-v6i3p118 This paper proposes TFX-MARL (Trusted Federated Ex-plainability for MARL), a governance-inspired framework for zero-shot policy transfer across silos using trust metric-based federated learning (FL) and explainability controls. TFX-MARL contributes: ... |
| [v16699] | Synaptic Failure is a Flat Minima Optimizer https://www.semanticscholar.org/paper/73f11953bef1953f5d530df702a68bf403de34b7 In addition to the effect on overfitting, we explore NormOut's impact on adversarial robustness against a suite of white and black-box attacks. Intriguingly, we find that some variants of NormOut produce extreme gradient masking without obfuscation. ... |
| [v16720] | On this day in tech history: In 1956, MIT researchers quietly tested the "Summer Vision Project precursor" camera rig, a hacked-together analog scanner used only in internal demos. https://aibreakfast.beehiiv.com/p/anthropic-to-go-public They handle multi-step reasoning, sub-task decomposition, and adapt to context dynamically. NotebookLM now supports prompts up to 10,000 characters, enabling detailed AI personas for work, education, and research. iOS features for infographics and s... |
| [v16772] | ONG: One-Shot NMF-based Gradient Masking for Efficient Model Sparsification https://arxiv.org/abs/2508.12891 Abstract: Deep Neural Networks (DNNs) have achieved remarkable success but their large size poses deployment challenges. While various pruning techniques exist, many involve complex iterative processes, specialized criteria, or struggle to maintain s... |
| [v16776] | Bayesian Mediation Analysis with an Application to Explore Racial Disparities in the Diagnostic Age of Breast Cancer https://doi.org/10.3390/stats7020022 Firstly, it allows us to make inferences on mediation effects based on the posterior distributions of parameters, eliminating the need for bootstrap sampling as we can directly obtain variances of estimates. Secondly, parameters are considered random... |
| [v16803] | Objective: The objective of the study is to build models for early prediction of risk for developing multiple organ dysfunction (MOD) in pediatric intensive care unit (PICU) patients. https://www.frontiersin.org/journals/pediatrics/articles/10.3389/fped.2021.711104/full All models were built in R (version 3.5.3) using the open source CRAN packages: xgboost (26), ranger (27), mboost (32), and glmnet (24), respectively, for the above methods. The choice of the above four methods was driven by the amount of available d... |
| [v16833] | Phase-Associative Memory: Sequence Modeling in Complex Hilbert Space https://arxiv.org/abs/2604.05030 However, their adoption in domains that require guaranteed reliability has been hindered by persistent difficulties, most prominently hallucination and susceptibility to prompt injection , which have resisted solution despite substantial engineering... |
| [v16836] | ZeroGrad : Mitigating and Explaining Catastrophic Overfitting in FGSM Adversarial Training https://arxiv.org/abs/2103.15476 Its goal is to evaluate robustness of models in a reliable manner and identify the defenses that give a wrong impression of robustness. Many earlier proposed defenses resulted in much lower robust accuracy compared to other common attacks that are us... |
| [v16866] | Austin is PI for new DoD Minerva Research... https://cee.umd.edu/news/story/austin-is-pi-for-new-dod-minerva-research-initiative-project Results will represent a significant step toward interoperable, reconfigurable, and traceable system capabilities. "Our research will provide the ability to imagine and explore alternative institutional designs," Austin said. ""This includes organiz... |
| [v16891] | Decision Transparency Enhancement And Integration Of User Feedback And Control Of Artificial Intelligence Outputs https://ppubs.uspto.gov/pubwebapp/external.html?q=(20260127199).pn The disclosed subject matter, in some embodiments thereof, relates to artificial intelligence explainability and customization and, more specifically, but not exclusively, to decision transparency enhancement and integration of user feedback and cont... |
| [v16904] | 2025: As organizations deploy millions of smart devices, the challenge of managing identity, access, and secure connectivity becomes mission-critical. https://shreyaswebmediasolutions.com/technology/securing-the-edge-how-idaas-supercharges-identity-management-in-aws-iot-core/ A Zero Trust model assumes no implicit trust - every device, user, or app must continuously prove its identity. When combined with AWS IoT Core, IDaaS enables this model by: Context-aware access (e.g., deny connections from unknown IPs or geo-zones)... |
| [v16996] | Novel Federated Graph Contrastive Learning for IoMT Security: Protecting Data Poisoning and Inference Attacks https://www.mdpi.com/2227-7390/13/15/2471 Both variants successfully reduced the number of communication rounds by almost 50% compared to traditional FedAvg, thereby confirming communication efficiency. However, the attention mechanisms need a lot of computing power, using function call grap... |
| [v17005] | The Geometric Canary: Predicting Steerability and Detecting Drift via Representational Stability https://arxiv.org/abs/2604.17698 Representation Engineering (Zou et al., 2023) and causal interventions (Meng et al., 2022;Geiger et al., 2024) rely on the Linear Representation Hypothesis (Park et al., 2023(Park et al., , 2025)), which posits that concepts are encoded as stable lin... |
| [v17029] | Anthropomorphism-based causal and responsibility attributions to robots https://doi.org/10.1038/s41598-023-39435-5 It is not always clear whether a human or robot was the cause of a failure in interactive situations. Nevertheless, a person will sometimes infer a cause and attribute responsibility to somebody or something for the failure, as is the case in the hum... |
| 1 | Home / Insights / Promise and Peril in the Age of Agentic AI: Navigating the New Security Landscape 2026-01-23 Research indicates that treating agents as privileged users requires robust identity governance, including multi-factor authentication adaptations and just-in-time provisioning mechanisms. 1.2.4 Agent Communication Poisoning In complex enterprise deployments, multiple agents will need to collaborate to accomplish sophisticated tasks. This inter-agent communication introduces vulnerabilities to poisoning attacks, where malicious actors inject false information into agent dialogues. Such attacks c... |
| 2 | LLM-TOC: LLM-Driven Theory-of-Mind Adversarial Curriculum for Multi-Agent Generalization 2026-03-07 To address these limitations, we propose LLM-TOC (LLM-Driven Theory-of-Mind Adversarial Curriculum), which casts generalization as a bi-level Stackelberg game: in the inner loop, a MARL agent (the follower) minimizes regret against a fixed population, while in the outer loop, an LLM serves as a semantic oracle that generates executable adversarial or cooperative strategies in a Turing-complete code space to maximize the agent's regret. To cope with the absence of gradients in discrete code gener... |
| 3 | Feature Distillation With Guided Adversarial Contrastive Learning 2020-09-20 Due to gradient masking, defensive distillation improves the robustness of the student model under a certain attack. (2020)... |
| 4 | user@alignchronicles : ~/posts $ cat scrutinizing-saliency-based-image-cropping. 2026-04-15 As it is evident in these example images, even the cropped image seems fair , the cropping has in fact, masked the differential saliency that the machine learning model associates with the different constituent faces in the image and some of these nuanced facets of biased ugliness are obfuscated in the finally rendered image. On the saliency model we used for the gradio app Given that both twitter's saliency-estimation model and the cropping policy are not in the public domain, we used a similar... |
| 5 | Management and Organization Review (1) 2026-02-09 We identify an accelerator by performing counterfactual expenditure increments on a particular policy issue while leaving the remaining ones with their original budgets. Then, a policy can be conceived as a systemic bottleneck when the removal of funding indirectly hinders the performance of other policy issues.... |
| 6 | Adaptive Defense Orchestration for RAG: A Sentinel-Strategist Architecture against Multi-Vector Attacks 2026-04-21 Attack and benchmark-focused work either targets a single class of adversary, such as membership inference against RAG , or concentrates on knowledge-base corruption and prompt-injection style poisoning without modeling privacy leakage . To the best of our knowledge, we are not aware of prior empirical work that simultaneously (i) evaluates RAG under concurrent multi-vector threats, specifically membership inference and data poisoning in our empirical study, while architecturally designing for c... |
| 7 | Too Polite to Disagree: Understanding Sycophancy Propagation in Multi-Agent Systems 2026-04-02 In multi-agent settings, Du et al. (2024) show that LLM instances debating over rounds can improve reasoning and reduce hallucinations.Estornell & Liu (2024) formalize this theoretically and show that similar model capabilities can cause convergence to incorrect majority opinions, proposing interventions such as misconception-refutation.ReConcile (Chen et al., 2024) improves consensus via confidence-weighted voting, and ConsensAgent (Pitre et al., 2025) targets copying via prompt refinement.Howe... |
| 8 | D-REX: A Benchmark for Detecting Deceptive Reasoning in Large Language Models 2025-09-21 D-REX was constructed through a competitive red-teaming exercise where participants crafted adversarial system prompts to induce such deceptive behaviors. Each sample in D-REX contains the adversarial system prompt, an end-user's test query, the model's seemingly innocuous response, and, crucially, the model's internal chain-of-thought, which reveals the underlying malicious intent.... |
| 9 | 3D-VCD: Hallucination Mitigation in 3D-LLM Embodied Agents through Visual Contrastive Decoding 2026-04-12 Abstract: Large multimodal models are increasingly used as the reasoning core of embodied agents operating in 3D environments, yet they remain prone to hallucinations that can produce unsafe and ungrounded decisions. Existing inference-time hallucination mitigation methods largely target 2D vision-language settings and do not transfer to embodied 3D reasoning, where failures arise from object presence, spatial layout, and geometric grounding rather than pixel-level inconsistencies.... |
| 10 | Systems-Level Attack Surface of Edge Agent Deployments on IoT 2026-02-25 All inter-agent communication uses MQTT pub/sub on the Mac mini broker (port 1883, Tailscale mesh only; no public exposure).Agents publish to topic-structured channels using a JSON envelope carrying sender ID, message type, microsecond timestamp, correlation ID, and payload.The NUC bridges MQTT to Home Assistant's REST API for IoT device control.Model inference calls traverse WAN to cloud providers; all operational IoT traffic remains mesh-local. This design makes MQTT the sole coordination plan... |
| 11 | HanoiWorld : A Joint Embedding Predictive Architecture BasedWorld Model for Autonomous Vehicle Controller 2026-01-03 Based on these aforementioned works, this result argue that world-model designing can be potential benefit from the high-quality self-supervised learning embedding from pretrained encoder as V-JEPA 2 and combine with the usage of long-term planner which can reduce and minimalize the cost of inference while remaining accuracy, and tunable model driving quality. The contribution of this studies include 4 keys essential contributions as follow: A unified perspective on world-model design for autono... |
| 12 | Counterfactual explanations and adversarial attacks have a related goal: flipping output labels with minimal perturbations regardless of their characteristics. 2026-03-17 Counterfactual explanations and adversarial attacks have a related goal: flipping output labels with minimal perturbations regardless of their characteristics. Yet, adversarial attacks cannot be used directly in a counterfactual explanation perspective, as such perturbations are perceived as noise and not as actionable and understandable image modifications.... |
| 13 | In an era where data privacy concerns increasingly shape public acceptance of digital health technologies, a new study states that advanced AI does not have to come at the cost of patient confidentia 2026-02-17 Errors tend to occur in borderline cases, such as early-stage disease or intermediate biomarker values, highlighting the importance of integrating AI outputs with clinical decision support rather than using them in isolation. This reinforces the view that federated AI systems should augment, not replace, human judgment in healthcare. The authors note that future work should incorporate explainability techniques, real-world clinical validation, and robust defenses against adversarial attacks to s... |
| 14 | Security-Aware Sensor Fusion with MATE: the Multi-Agent Trust Estimator 2025-11-18 The security-aware sensor fusion both detects misbehaving agents and recovers accurate SA under adversarial manipulation. Trust estimation is a two-step hidden Markov model (HMM). The first step is to propagate the estimate forward in time. The second step is to update the estimate with measurements. Since there is no sensor providing direct measurements of trust (unlike e.g., GPS providing position), we design a novel method of mapping real perception-oriented sensor data to trust pseudomeasure... |
| 15 | Boosting Value Decomposition via Unit-Wise Attentive State Representation for Cooperative Multi-Agent Reinforcement Learning 2025-12-31 For the problems of non-stationarity and partial observability, an appealing paradigm is Centralized Training and Decentralized Execution (CTDE).... |
| 16 | The Architectural Evolution of Intelligence: A Formal Taxonomy of the AI Technology Stack 2026-05-10 The enterprise utility is significant: Knowledge Graphs constructed via RDF/OWL provide the structured "world model" that prevents higher-level agents from confabulating organizational hierarchies, regulatory relationships, or product taxonomy structures. Grounding a generative model against a formally specified ontology is the primary architectural defense against hallucination-induced operational failure. 2.4 Search Algorithms, Heuristics, and Combinatorial Optimization Operational enterprise ... |
| 17 | by Erik Jenner, Viktor Rehnberg, Oliver Daniels 2026-03-11 Better MAD proxies for scheming/deceptive alignment: As mentioned before, backdoor detection has some similarities to detecting a treacherous turn. But in data poisoning backdoor attacks (and for natural mechanism distinction), the model is explicitly trained to exhibit bad behavior. In contrast, the main worry for a scheming model is that it would exhibit bad behavior "zero-shot." This might affect which MAD methods are applicable. For example, finetuning on trusted data is a decent backdoor de... |
| 18 | InsightSwarm: A Multi-Agent Adversarial Framework for Automated Fact-Checking with Real-Time Source Verification, Human-in-the-Loop Oversight, and Adaptive Confidence Calibration 2026-04-29 InsightSwarm: A Multi-Agent Adversarial Framework for Automated Fact-Checking with Real-Time Source Verification, Human-in-the-Loop Oversight, and Adaptive Confidence Calibration --- FactChecker pipeline that independently fetches and validates every cited URL, reducing source hallucination to below 3 percent; (3) Human-in-the-Loop (HITL) intervention via LangGraph interrupt semantics enabling mid-pipeline human source correction through a live React panel; (4) adaptive confidence calibration us... |
| 19 | Differential privacy has become the gold standard for protecting individual data in analytics and machine learning, but it still relies on outdated assumptions about how people trust one another. 2026-01-24 By tailoring privacy guarantees to each user's local trust environment, TGDP can offer higher utility than local DP while maintaining more realistic privacy boundaries than central DP. It reflects a philosophical shift as much as a technical one: from privacy as a global policy to privacy as a networked, context-aware contract. How Trust Affects Accuracy In TGDP, privacy is tied to trust, but so is performance. The more people you trust (and who trust each other), the more accurately you can com... |
| 20 | The Artificial Intelligence in Social Media Market grew from USD 3.14 billion in 2025 to USD 3.90 billion in 2026. 2026-04-14 In the Americas, rapid adoption of cloud-native services, a vibrant creator economy, and well-established advertising ecosystems favor experimentation with generative content and predictive targeting, while regulatory debates and privacy concerns push firms to prioritize transparency and consent mechanisms. Europe, Middle East & Africa presents a mosaic of regulatory regimes and infrastructure capacities, where firms must navigate stringent data protection requirements, local content norms, and ... |
| 21 | Aetheria: A multimodal interpretable content safety framework based on multi-agent debate and collaboration 2025-12-01 More importantly, these monolithic systems inevitably suffer from single-model biases and hallucinations . They often demonstrate insufficient capability in identifying implicit risks that require deep reasoning and diverse cultural contextual knowledge , failing to meet the dual requirements of comprehensiveness and interpretability . As illustrated in table 1, existing paradigms often fail to simultaneously satisfy the critical requirements of implicit risk detection, interpretability, and mul... |
| 22 | Understanding the Information Propagation Effects of Communication Topologies in LLM-based Multi-Agent Systems 2025-05-28 Motivated by our Insight, EIB-LEARNER balances the error-insight trade-off by co-training two complementary graph neural network (GNN) simulators to simulate the error suppression and insight propagation given a specific query (Section 4.1), and then adaptively blending their learned inter-agent coefficients to construct robust topologies (Section 4.2).The overall pipeline of EIB-LEARNER is shown in Figure 3. GNN-based Propagation Simulators To balance error suppression and insight propagation i... |
| 23 | Deliberative Alignment: Reasoning Enables Safer Language Models 2024-12-19 Deliberative Alignment: Reasoning Enables Safer Language Models --- Alternatively, an AI could remain committed to its human-assigned terminal goal but, in the process, pursue instrumental goals like self-preservation, resource acquisition, or enhancing its cognitive abilities , . These power-seeking tendencies could lead to harmful or unintended consequences. And as models gain more intelligence and autonomy, the scale of potential harm from misalignment increases dramatically, with the risk of... |
| 24 | Systems and Methods for Protecting Machine Learning (ML) Units, Artificial Intelligence (AI) Units, Large Language Model (LLM) Units, Deep Learning (DL) Units, and Reinforcement Learning (RL) Units 2026-01-14 Systems and Methods for Protecting Machine Learning (ML) Units, Artificial Intelligence (AI) Units, Large Language Model (LLM) Units, Deep Learning (DL) Units, and Reinforcement Learning (RL) Units --- wherein the Explainability Module is further configured to enable consent management and provenance capture.... |
| 25 | Optimization under Attack: Resilience, Vulnerability, and the Path to Collapse 2025-02-08 Notable advancements include extensions of consensus-based protocols by Sundaram et al. and Kuwaranancharoen et al. , which address adversarial threats in convex optimization. Su et al. enhance these methods with decentralized architectures and explore adversarial influence on global objectives. However, these approaches assume adversary agents have full knowledge of the network topology and the private functions of all agents. This coordination among adversaries compromises the privacy of the a... |
| 26 | A Unified Framework for Evaluating and Enhancing the Transparency of Explainable AI Methods via Perturbation-Gradient Consensus Attribution 2024-12-04 Perturbation-based methods achieve high fidelity by directly querying the model, while gradient-based methods achieve high robustness through deterministic gradient computation. By fusing both paradigms through consensus amplification, PGCA inherits the advantages of each while mitigating their individual weaknesses. The complete algorithmic specification is provided in Algorithm 1, and each stage is analyzed below. Stage 1 generates a perturbation importance map using an 8 8 grid (64 cells), te... |
| 27 | TxRay: Agentic Postmortem of Live Blockchain Attacks 2026-01-31 The following key takeaways summarize the main challenges: (i) Filling information gaps under partial observability.... |
| 28 | Interpreting Agentic Systems: Beyond Model Explanations to System-Level Accountability 2026-01-22 These limitations make LIME's explanations fragmentary and potentially unreliable for understanding an agentic system's behavior. Attention/Saliency Maps: For models like transformers, one might attempt to use attention weights or gradient-based saliency as explanations (e.g. highlighting which words or state elements an agent "focused" on). This, too, has limited utility in agentic systems. In a multi-agent LLM system, an agent's policy might not even expose attention weights to the end-user, a... |
| 29 | Tacit mechanism: Bridging pre-training of individuality to multi-agent adversarial coordination 2026-01-31 For pre-training the tacit behaviors, we develop a pattern mechanism and a tacit mechanism to integrate spatial relationships among agents, which dynamically guide agents' actions to gain spatial advantages for coordination. In the subsequent centralized adversarial training phase, we utilize the pre-trained network to enhance the formation of advantageous spatial positioning, achieving more efficient learning performance.... |
| 30 | Global Prediction of Dengue Incidence Using an Explainable Artificial Intelligence - Driven ConvLSTM Integrating Environmental, Health, and Socio - Economic Determinants 2026-04-05 ... y^i-yi|,R2=1- i=1n(y^i-yi) in(y^i-y ) Where, n denotes the number of observations and p the number of predictors. 2.3.6 Feature Contribution and Sensitivity Analyses Using SHAP SHapley Additive exPlanations (SHAP) and permutation - based importance were used to quantify predictor contributions. SHAP values for feature i are: i= S F{i}|S|!(|F|-|S|-1)!|F|[fs {i}(XS {i})-fs(xs)] Where, F is the set of all features, S is a subset of features excluding i, fs(xs)denotes the model prediction using ... |
| 31 | The remarkable growth and adoption of machine learning models have brought along an uncomfortable reality: these systems can be manipulated, deceived, and corrupted by adversarial inputs. 2026-04-18 Another line of defenses includes detection mechanisms - identifying when an input is suspiciously adversarial. In practice, though, detection often lags behind sophisticated new attacks. For model poisoning, robust aggregation rules can mitigate malicious updates in federated learning scenarios (where partial updates from multiple participants are combined).... |
| 32 | Unified World Models: Memory-Augmented Planning and Foresight for Visual Navigation 2025-10-08 Humans naturally excel at such imaginative reasoning, routinely performing mental simulations to plan routes effectively through both familiar and novel scenarios Bar et al. (2025). Despite rapid progress in visual navigation, existing approaches remain constrained by fundamental limitations (Figs. 1). (a) Direct policy methods (e.g., GNM Shah et al. (2022), VINT Shah et al. (2023), NoMaD Sridhar et al. (2024)) map observations directly to action sequences. Although effective within familiar dis... |
| 33 | What Is an AI-Enabled Cyber-Attack? 2026-04-18 Since ChatGPT's launch, phishing volume has surged by 4,151%, demonstrating how AI removes the bottlenecks that once limited attack campaigns. Precision targeting that actually works: AI-generated phishing emails achieve a 54% success rate compared to just 12% for traditional attacks. Attackers can now scrape social media profiles, corporate websites, and public records to create hyper-personalised messages that reference recent purchases, mutual contacts, or company-specific terminology. Democr... |
| 34 | LLM-TOC: LLM-Driven Theory-of-Mind Adversarial Curriculum for Multi-Agent Generalization 2026-03-07 To address these limitations, we propose LLM-TOC (LLM-Driven Theory-of-Mind Adversarial Curriculum), which casts generalization as a bi-level Stackelberg game: in the inner loop, a MARL agent (the follower) minimizes regret against a fixed population, while in the outer loop, an LLM serves as a semantic oracle that generates executable adversarial or cooperative strategies in a Turing-complete code space to maximize the agent's regret.... |
| 35 | Reinforcement Learning (RL) has emerged as a pivotal and transformative subset of machine learning, enabling autonomous agents to acquire optimal behaviors and decision-making policies through iterat 2026-02-19 The integration of RL with deep neural networks has particularly revolutionized its practical applicability, enabling agents to process high-dimensional sensory data and achieve superhuman performance in domains ranging from strategic games and robotic control to autonomous navigation and precision healthcare. However, the widespread and responsible deployment of RL systems hinges on diligently addressing several critical challenges. The inherent demand for vast amounts of interaction data neces... |
| 36 | VEM: Environment-Free Exploration for Training GUI Agent with Value Environment Model 2025-02-25 We now provide a more advanced argument showing that if Q θ approximates Q * , i.e., the optimal value model, on the support of D, then the learned policy π can achieve near-optimal returns. In addition, we introduce distribution shift considerations and demonstrate how coverage of D influences policy quality. Offline Coverage and Value Approximation. We introduce two conditions which bounds the suboptimality gap relative to the optimal policy π * : Coverage Definition. For a policy π, define th... |
| 37 | Second Order Optimization for Adversarial Robustness and Interpretability 2020-09-09 The relationship between adversarial robustness and saliency map interpretability was recently studied in (Etmann et al. 2019) but experiments were based on gradient regularization. Furthermore, recent works Ilyas et al. 2019) claim that existence of adversarial examples are due to standard training methods that rely on highly predictive but non-robust features, and make connections between robustness and explainability. In this paper, we propose a quadratic-approximation of adversarial attacks ... |
| 38 | Distributed Nonlinear Control of Networked Two-Wheeled Robots under Adversarial Interactions 2026-04-04 ... goal of fully distributed implementation and increase vulnerability to coordinated attacks. Addressing resilience for nonlinear, nonholonomic multi-agent systems under adversarial information exchange therefore remains an open and practically relevant problem . Other secure multi-agent coordination methods use homomorphic encryption techniques combined with distributed control approaches to ensure secure computation of distributed control through third-party cloud services . In this paper, w... |
| 39 | The impact of machine learning uncertainty on the robustness of counterfactual explanations 2026-04-30 Through experiments on synthetic and real-world tabular datasets, we show that counterfactual explanations are highly sensitive to model uncertainty.In particular, we find that even small reductions in model accuracy -caused by increased noise or limited data -can lead to large variations in the generated counterfactuals on average and on individual instances.These findings underscore the need for uncertainty-aware explanation methods in domains such as finance and the social sciences. Introduct... |
| 40 | Modeling what Matters: Emergent Abstraction In Reinforcement Learning - Robotics Institute Carnegie Mellon University 2026-04-17 Modeling what Matters: Emergent Abstraction In Reinforcement Learning - Robotics Institute Carnegie Mellon University Modeling what Matters: Emergent Abstraction In Reinforcement Learning 2025-12-12 15:00:002025-12-12 16:30:00 Benjamin (Ben) Freed PhD Student Robotics Institute, Abstract: Real-world decision-making is rife with partial observability, long horizons, and complex multi-agent interactions. This thesis argues that abstraction - forming simplified representations of the task that reta... |
| 41 | Constrained Black-Box Attacks Against Multi-Agent Reinforcement Learning 2025-12-31 In this paper, we investigate new vulnerabilities under more realistic and constrained conditions, assuming an adversary can only collect and perturb the observations of deployed agents.We also consider scenarios where the adversary has no access at all.We propose simple yet highly effective algorithms for generating adversarial perturbations designed to misalign how victim agents perceive their environment.... |
| 42 | SlimComm: Doppler-Guided Sparse Queries for Bandwidth-Efficient Cooperative 3-D Perception 2025-08-17 An agent becomes a collaborator whenever at least one query lands on a BEV cell whose warped foreground density exceeds the communication threshold: max where (, ) are BEV grid indices. The test is performed only at the finest scale =0, whose higher resolution captures the most detailed occupancy information. Halo-enriched Sparse Feature Encoding. Most existing methods [6,16,26,29] perform early-stage projection: they first transform every CAV's point cloud into the ego frame and then learn all ... |
| 43 | Shanxi Normal University, Taiyuan, China 2026-01-13 Abstract:Multi-agent reinforcement learning typically employs a centralized training-decentralized execution (CTDE) framework to alleviate the non-stationarity in environment. However, the partial observability during execution may lead to cumulative gap errors gathered by agents, impairing the training of effective collaborative policies.... |
| 44 | GH Research PLC: EXHIBIT 99.2 (EX-99.2) 2026-05-13 In November 2025, we submitted a complete response to the clinical hold and in December 2025, the hold was lifted by the FDA. In parallel, we are conducting the Phase 1 healthy volunteer clinical pharmacology trial (GH001-HV-106) using our proprietary device in the United Kingdom. GH002 is our second mebufotenin product candidate, formulated for administration via a proprietary intravenous injection approach. We have completed a randomized, double-blind, placebo-controlled, dose-ranging clinical... |
| 45 | You know the saying: it takes all sorts? 2026-03-15 Root cause analysis usually identifies one or a small number of factors, and attributes blame. Mess mapping reveals the systemic nature of such failures, and avoids the fundamental attribution error: blaming someone while ignoring the context in which they worked. The red team This well-known adversarial approach has applications beyond the military and cybersecurity.... |
| 46 | Robust Coordination Under Misaligned Communication via Power Regularization 2024-04-08 Within this framework, communication is understood through the perspectives of information theory and control, defined as the exchange of information between agents via an established channel, typically employed to facilitate coordination. In contrast, Cooperative Multi-Agent Reinforcement Learning (CoMARL) generally emphasizes parameter-sharing, optimizing team training efficiency, and developing cooperative mechanisms to address collective challenges. While many CoMARL algorithms leverage para... |
| 47 | ICLR 2026 produced a failure playbook for multi-agent systems. 2026-04-18 The mundane, reproducible, expensive kind of failures that happen when you deploy these systems in production and watch your latency quadruple while your error rate climbs. The papers cluster into three failure modes: agents that talk too much, agents that coordinate too slowly, and agents that break each other in cascades. Each cluster comes with proposed fixes, and the fixes are where the research gets interesting. But the failures come first, because the field has been building multi-agent sy... |
| 48 | Every production database needs a plan for when things go wrong. 2026-04-23 Fraud detection and anomaly monitoring systems that rely on similarity search to flag suspicious activity - a gap in coverage creates a window of vulnerability. Autonomous agent systems that use vector stores for memory and tool retrieval - agents fail or loop without their knowledge base. If you're evaluating vector databases for any of these use cases, high availability isn't a nice-to-have feature to check later. It should be one of the first things you look at. What Does Production-Grade HA ... |
| 49 | Customer data ethics and transparency technology has emerged as a critical infrastructure requirement for marketing organizations navigating an era where consumer data practices face unprecedented s 2026-04-17 Fairness constraints can be applied during algorithm training to ensure that model outputs maintain equitable treatment across defined groups while preserving overall marketing effectiveness. Ongoing monitoring systems continuously evaluate deployed algorithms for emerging bias patterns that may develop as customer populations, market conditions, or data distributions evolve after initial model deployment. Explainability tools provide human-interpretable explanations of why specific algorithmic ... |
| 50 | Methods For Prediction Of Neutronics Parameters Using Deep Learning 2024-02-21 Methods For Prediction Of Neutronics Parameters Using Deep Learning --- Therefore, the data-driven model - LatticeNet, in this case - is able to combine the accuracy strengths of a high-fidelity solver (MPACT) with the computational strengths of low-fidelity nodal methods. The primary benefit that both of these methods have, which LatticeNet does not, is explainability; as far as the authors are aware, there are no techniques for decoding "why" a neural network gives the answer it does. Current ... |
| 51 | Enhancing Hallucination Detection in Large Language Models through a Dual-Position Debate Multi-Agent Framework 2025-11-09 Enhancing Hallucination Detection in Large Language Models through a Dual-Position Debate Multi-Agent Framework --- This paper introduces a novel Dual-Position Debate DPD framework designed to enhance the veracity of LLM-generated content and mitigate hallucinations.... |
| 52 | Towards Detecting LLMs Hallucination via Markov Chain-based Multi-agent Debate Framework 2024-06-06 To overcome these limitations, we propose a Markov Chain-based multi-agent debate verification framework to enhance hallucination detection accuracy in concise claims. Our method integrates the fact-checking process, including claim detection, evidence retrieval, and multi-agent verification.... |
| 53 | Sync or Sink: Bounds on Algorithmic Collective Action with Noise and Multiple Groups 2025-12-31 Because they are targeting two different classes, the suboptimality gap may also be large.They also find a case where two collectives, with different target classes and different character usage, still sinks both of their success rates.This can also be explained by the cross-signal overlap -if these character modifications look sufficiently "close" to each other, this term may be large and cause conflicts.Figure 5: Impact of noise (Random-subset) on the feature-only strategy.Compared to the feat... |
| 54 | Credit Assignment with Meta-Policy Gradient for Multi-Agent Reinforcement Learning 2021-02-23 Reward decomposition is a critical problem in centralized training with decentralized execution~(CTDE) paradigm for multi-agent reinforcement learning. (2021)... |
| 55 | This paper demonstrates how reinforcement learning can explain two puzzling empirical patterns in household consumption behavior during economic downturns. 2026-04-21 As a first step towards model-free Bayes optimality, we introduce the Bayesian exploration network (BEN) which uses normalising flows to model both the aleatoric uncertainty (via density estimation) and epistemic uncertainty (via variational inference) in the Bellman operator. In the limit of complete optimisation, BEN learns true Bayes-optimal policies, but like in variational expectation-maximisation, partial optimisation renders our approach tractable. Empirical results demonstrate that BEN c... |
| 56 | FLARE: Adaptive Multi-Dimensional Reputation for Robust Client Reliability in Federated Learning 2026-05-13 Abstract: Federated learning (FL) enables collaborative model training while preserving data privacy. However, it remains vulnerable to malicious clients who compromise model integrity through Byzantine attacks, data poisoning, or adaptive adversarial behaviors. Existing defense mechanisms rely on static thresholds and binary classification, failing to adapt to evolving client behaviors in real-world deployments. We propose FLARE, an adaptive reputation-based framework that transforms client rel... |
| 57 | It's Wednesday, February 25, 2026, and here are the top tech stories making waves today. 2026-03-09 For startups building "AI for gov," it's a signal that the bar is rising: winning won't just be about model quality, but about compliance, integration, and trust frameworks. Why It Matters: Government adoption of frontier AI in classified workflows can reshape the competitive landscape for enterprise AI - and accelerate regulation expectations. Amazon's AI coding tool backlash shows the limits of "blame the human" narratives The Register describes internal turbulence around Amazon's AI coding ef... |
| 58 | Enhancing Heterogeneous Multi-Agent Cooperation in Decentralized MARL via GNN-driven Intrinsic Rewards 2024-08-11 We additionally compare with the state-of-the-art MARL baseline, IPPO (Independent Proximal Policy Optimization), which is applicable in decentralized training settings for heterogeneous agents under partial observability similar to HetGPPO. Unlike the two centralized critic-based heterogeneous MARL approaches discussed in the 'Related Works' section or widely used algorithms such as MADDPG , MAPPO , COMA , etc., these baselines along with CoHet address the more challenging problem of not relyin... |
| 59 | by Esben Kran, HaydnBelfield, Apart Research 2026-04-22 Curious to see more generality testing for the inverse scaling. See the dataset generation code, the graph plotting code, and the report. By Clement Dumas, Charbel-Raphael Segerie, Liam Imadache Abstract: Neural Trojans are one of the most common adversarial attacks out there. Even though they have been extensively studied in computer vision, they can also easily target LLMs and transformer based architecture. Researchers have designed multiple ways of poisoning datasets in order to create a bac... |
| 60 | Is AI secretly learning from you? The unseen power of federated learning 2025-04-01 Federated learning design: How federated learning can be applied in decentralized environments. Implementation challenges: Combating data traffic jams, delay issues, and security risks. Advanced model aggregation: How to combine many devices' contributions without compromising accuracy. Security measures: How to prevent attacks, data poisoning, and adversarial risks.... |
| 61 | Towards desiderata-driven design of visual counterfactual explainers 2026-05-07 This can be e.g. the inclusion or removal of object parts, but also more intricate changes in image quality or color, that may not be accessible with other explanation techniques such as feature attribution.Another advantage of counterfactuals is that they are inherently actionable, e.g.together with a human in the loop, counterfactuals provide an implicit data augmentation scheme that can serve to address a model's missing invariances or reliance on spurious correlations .Mathematically, the se... |
| 62 | ZTFed-MAS2S: A Zero-Trust Federated Learning Framework with Verifiable Privacy and Trust-Aware Aggregation for Wind Power Data Imputation 2025-08-23 1) The ZTFed framework integrates verifiable Differential Privacy with Non-Interactive Zero-Knowledge Proofs (DP-NIZK) and a Confidentiality and Integrity Verification (CIV) mechanism to enable verifiable privacy preservation and secure, integrity-assured model transmission. In addition, it employs a Dynamic Trust-Aware Aggregation (DTAA) mechanism to enhance resilience against anomalous clients and incorporates sparsity-and quantization-based compression to reduce communication overhead. 2) The... |
| 63 | Misalignment in Multi-Agent Systems (MAS) is frequently treated as a technical failure. 2025-12-31 Just as perception shifts in the illusion, MAS frameworks can be framed differently depending on theoretical or empirical perspectives, leading to inconsistent definitions of coordination and cooperation.In complex or uncertain environments, incomplete knowledge and partial observability further blur the distinction between coordinating tasks and cooperating for collective benefit, thereby amplifying the reach of the Misalignment Mosaic.While the Rabbit-Duck illusion broadly represents perceptua... |
| 64 | Towards Detecting LLMs Hallucination via Markov Chain-based Multi-agent Debate Framework 2025-04-05 To overcome these limitations, we propose a Markov Chain-based multi-agent debate verification framework to enhance hallucination detection accuracy in concise claims.... |
| 65 | The Essence of Balance for Self-Improving Agents in Vision-and-Language Navigation 2026-04-20 On the one hand, the agent benefits from behavioral diversity-maintaining multiple plausible latent hypotheses for the next action under linguistic ambiguity and partial observability.On the other hand, self-improvement from policy-induced trajectories requires learning stability, so that updates remain consistent enough to accumulate progress across iterations.This creates an inherent tension: increasing diversity can uncover better hypotheses under ambiguity, but may introduce inefficient expl... |
| 66 | In the case for CoT unfaithfulness is overstated, @nostalgebraist pointed out that reading the chain-of-thought (CoT) reasoning of models is neglected as an interpretability technique. 2026-04-19 We can reduce the risk of steganography by forcing the agent to decompose its task into subtasks, eliminating unnecessary added context that could be used to pass on steganographic messages. Here's a more concrete description: consider a "tree" of agents. The top-level agent receives the user's query and can think about how to solve it, but it has a very limited token budget for its thoughts. However, it can get more thinking done by delegating to other AI instances (either of itself or of a sma... |
| 67 | LLM observability is the practice of tracing, measuring, and understanding how large language model applications behave in production - connecting inputs, outputs, and internal steps to explain why a 2026-03-09 With LLM observability, you trace the failing request, discover that the vector store returned irrelevant chunks due to an embedding model update, and pinpoint that the prompt template lacked grounding instructions. You fix the retrieval step - not the model. Cost Attribution Across Multi-Agent Workflows An engineering team runs five agents: a code reviewer, a security scanner, a test generator, a documentation writer, and an issue triager. Monthly LLM costs hit $40,000 and the VP of Engineering... |
| 68 | grag-system added to PyPI 2026-05-12 Production-grade Graph RAG system combining knowledge graph reasoning, vector similarity search, reinforcement learning self-improvement, and explainable AI all in a single pip install. ... ... parse("What deep learning frameworks did Google create in 2017?")# parsed.intent "entity_info"# parsed.entities # parsed.constraints {"year": 2017, "domain": "ml"} Stage 2 Hybrid Retrieval Combines vector similarity with knowledge-graph-neighbor boosting. fromgrag.retrieval.hybrid_retrieverimportHybridRet... |
| 69 | UniC-RAG: Universal Knowledge Corruption Attacks to Retrieval-Augmented Generation 2025-08-25 We conduct systematic evaluations of UniC-RAG on 4 question-answering datasets: Natural Question (NQ) , HotpotQA , MS-MARCO , and a dataset (called Wikipedia) we constructed to simulate real-world RAG systems using Wikipedia dump .We also conduct a comprehensive ablation study containing 4 RAG retrievers, 7 LLMs varying in architectures and scales (e.g., Llama3 , GPT-4o ), and different hyperparameters of UniC-RAG.We adopt Retrieval Success Rate (RSR) and Attack Success Rate (ASR) as evaluation ... |
| 70 | The integration of autonomous decision-making frameworks within Web3 ecosystems represents a profound and transformative advancement in decentralized technologies. 2026-02-08 As the number of agents and the complexity of their tasks increase, ensuring efficient computation for AI models (especially on-chain inference), secure decentralized off-chain computation, and effective coordination mechanisms becomes paramount. Solutions may involve specialized Layer 2 scaling solutions designed for agent-centric computation, parallel processing architectures, and advanced multi-agent reinforcement learning (MARL) techniques to optimize cooperative behaviors. Security and Robu... |
| 72 | CoBel-World: Harnessing LLM Reasoning to Build a Collaborative Belief World for Optimizing Embodied Multi-Agent Collaboration 2025-09-25 CoBel-World: Harnessing LLM Reasoning to Build a Collaborative Belief World for Optimizing Embodied Multi-Agent Collaboration --- However, these approaches typically rely on fixed communication protocols, such as tep-by-step message generation (Zhang et al., 2023), eventdriven multi-round discussion (Liu et al., 2024b), or dense discussion (Guo et al., 2024), leading to excessive communication overhead and poor scalability under partial observability. In contrast, our work introduces a belief-dr... |
| 73 | Targeted Adversarial Poisoning Attack Against Robust Aggregation in Federated Learning for Smart Grids 2026-02-28 To counter these threats, secure aggregation rules have been implemented to reduce the impact of adversarial or malicious updates during training process. In this paper, we first propose a norm-based aggregation rule specifically designed to mitigate the effects of poisoning attacks within federated learning systems used for power quality classification.... |
| 74 | Sync or Sink: Bounds on Algorithmic Collective Action with Noise and Multiple Groups 2025-10-20 Sync or Sink: Bounds on Algorithmic Collective Action with Noise and Multiple Groups --- Because they are targeting two different classes, the suboptimality gap may also be large. They also find a case where two collectives, with different target classes and different character usage, still sinks both of their success rates. This can also be explained by the cross-signal overlap -if these character modifications look sufficiently "close" to each other, this term may be large and cause conflicts.... |
| 75 | Efficient and Trustworthy Block Propagation for Blockchain-Enabled Mobile Embodied AI Networks: A Graph Resfusion Approach 2025-01-25 When dealing with sensitive or critical information, malicious attacks can lead to severe consequences, such as information leakage, traffic accidents, or machine interaction failures. To mitigate these risks, the integration of blockchain technology is essential. The network layer, abstracted from the physical layer, presents the validator network in consortium blockchainsenabled MEANETs. The block propagation process is performed according to the mechanism detailed in Section III-A. Here, the ... |
| 76 | A Theory of Mind Approach as Test-Time Mitigation Against Emergent Adversarial Communication 2023-05-29 Explicitly, there are works on learning to communicate messages from CoMARL agents; however, non-cooperative agents have been shown to learn sabotage a cooperative team's performance through adversarial communication messages. To address this issue, we propose a technique which leverages local formulations of Theory-of-Mind (ToM) to distinguish exhibited cooperative behavior from non-cooperative behavior before accepting messages from any agent. We demonstrate the efficacy and feasibility of the... |
| 77 | Strategic Heterogeneous Multi-Agent Architecture for Cost-Effective Code Vulnerability Detection 2026-04-22 Du et al. show that having multiple LLMs debate improves factuality and reasoning, with agents correcting each other's errors through iterative rounds-a mechanism that directly inspires our adversarial verification loop. Liang et al. extend this to divergent thinking, finding that multi-agent debate elicits more diverse reasoning paths. CAMEL introduces role-playing communication protocols for multi-agent collaboration, demonstrating that specialized agent roles outperform generic prompting. The... |
| 78 | LLM Harms: A Taxonomy and Discussion 2025-12-04 LLM Harms: A Taxonomy and Discussion --- Redteaming plus rule-based "constitutional" fine-tuning cut jailbreak success by ~40 % on Llama 3-8B without crippling utility , yet toxic-speech filters still miss 7 % of non-English slurs . Third, governance levers are fragmentary: while the EU AI Act now imposes transparency and copyright duties on generalpurpose models , the U.S. leans on voluntary Risk-Management guidance and export-control tweaks targeting compute supply chains Federal Register. Ove... |
| 79 | Theoretical Guarantees for LT-TTD: A Unified Transformer-based Architecture for Two-Level Ranking Systems 2025-05-06 ... min θ L1 L L1 (θ L1 ) and min θ L2 L L2 (θ L2 )(3) independently.However, the optimal parameters θ * L1 for L1 may not lead to the best input for L2, and vice versa.An ideal system would jointly optimize: min θ L1 ,θ L2 L joint (θ L1 , θ L2 ) (4) Lemma 2 (Suboptimality of Disjoint Optimization).Let θ * L1 and θ * L2 be the optimal parameters when optimizing L L1 and L L2 independently, and let θ * joint be the optimal parameters when optimizing L joint .Then: L joint (θ * joint ) ... |
| 80 | Diffusion Counterfactuals for Image Regressors 2025-12-31 Adversarial Counterfactual Explanations (ACE) generate counterfactual images by optimizing adversarial perturbations in the image space while filtering high-frequency and out-of-distribution artifacts using a diffusion model. More specifically, consider L class (x, y) as a function that quantifies the match between a sample x and a class y, typically the cross-entropy loss, which we aim to minimize.Consider a filtering function F that constrains a counterfactual x ' to the data manifold of the t... |
| 81 | Amplification of formal method and fuzz testing to enable scalable assurance for communication system 2026-05-04 Numerous studies have shown vulnerabilities of the wireless communication links that allow intercepting, hijacking, or crashing UAVs via jamming, spoofing de-authentication, and false data injection. The cooperative nature of multi-UAV networks and the uncontrolled environment at low altitudes where they operate make it possible for malicious nodes to join and disrupt the routing protocols. While multi-node networks such as flying ad-hoc network (FANET) can extend the operational rage of UAVs, s... |
| 82 | Artificial Intelligence (AI) Automation Solutions Discovery Industry Disruptors / Game Changers Future Trends Tech Know How Insights into the Software Industry Business-IT Alignment Digital Twin Mac 2026-03-15 An RL agent is learning by making a mistake, but a mistake by an autonomous car or a heavy industrial robot can be catastrophic. Safe RL (SRL) techniques, which add hard constraints and risk metrics into the reward function, are a primary focus of the current research in this area. Data Efficiency and Sample Complexity: RL algorithms are sample-inefficient that require millions of data points (trials) to converge on a good policy. This means that they need highly accurate, large-scale simulators... |
| 83 | Optimal Robust Recourse with L p -Bounded Model Change 2025-12-31 Our Contributions and Results Our main goal is to understand the true price of recourse for more restricted adversarial model changes.In particular, we measure model changes by bounding the L p norm of the difference between initial and changed models, where p 1 but p = .We provide a new algorithm that provably computes the optimal robust recourse for generalized linear models for this type of model change. The key insight in the design of our algorithm is the observation that the optimal soluti... |
| 84 | Image Compression And Decoding, Video Compression And Decoding: Methods And Systems 2026-03-25 Note, during training the quantisation operation Q is not used, but we have to use it at inference time to obtain a strictly discrete latent. FIG. shows an example model architecture with side-information. The encoder network generates moments p and a together with the latent space y: the latent space is then normalised by these moments and trained against a normal prior distribution with mean zero and variance 1. When decoded, the latent space is denormalised using the same mean and variance. N... |
| 85 | Inherent Adversarial Robustness of Deep Spiking Neural Networks: Effects of Discrete Input Encoding and Non-linear Activations 2020-10-05 For example, an ensemble of defenses based on "gradient-masking" collapsed under the attack proposed in . Defensive distillation was broken by Carlini-Wagner method , . (2020)... |
| 86 | Revealing Vulnerabilities of Neural Networks in Parameter Learning and Defense Against Explanation-Aware Backdoors 2025-12-31 Rieger and Hansen devised an effective defense against adversarial attacks by combining multiple explanation methods, batting aside manipulation but possibly welcoming method-specific explanation.Lakkaraju et al. introduced a model training approach for producing resilient explanations, utilizing adversarial samples in training to discern discriminatory features.Gan et al. put forth MeTFA, a tool for enhancing explanation algorithm stabil-ity with theoretical guarantees, applicable to any featur... |
| 87 | Zero-Shot Policy Transfer in Multi-Agent Reinforcement Learning via Trusted Federated Explainability 2026-02-27 This paper proposes TFX-MARL (Trusted Federated Ex-plainability for MARL), a governance-inspired framework for zero-shot policy transfer across silos using trust metric-based federated learning (FL) and explainability controls. TFX-MARL contributes: (i) a trust metric that quantifies participant integrity and accountability using provenance, update consistency, local evaluation reliability, and safety-compliance signals; (ii) a trust-aware federated aggregation protocol that reduces poisoning ri... |
| 88 | Graph-Augmented Large Language Model Agents: Current Progress and Future Prospects 2025-07-28 Graph-Augmented Large Language Model Agents: Current Progress and Future Prospects --- Specifically, we categorize existing GLA methods by their primary functions in LLM agent systems, including planning, memory, and tool usage, and then analyze how graphs and graph learning algorithms contribute to each. For multi-agent systems, we further discuss how GLA solutions facilitate the orchestration, efficiency optimization, and trustworthiness of MAS. Finally, we highlight key future directions to a... |
| 89 | Adversarial Counterfactual Visual Explanations 2023-03-16 Yet, adversarial attacks cannot be used directly in a counterfactual explanation perspective, as such perturbations are perceived as noise and not as actionable and understandable image modifications. (2023)... |
| 90 | Traditional Chinese Medicine Can Be Seen as a Large Model Trained for Five Thousand Years 2026-03-09 AI's rapid progress has brought not only new tools but new epistemological shocks - shocks that help us reinterpret TCM. # 1. Large models challenge reductionism Modern science relies on "break down understand predict." But large models show that complex abilities can emerge from massive correlations without explicit causal modeling. Effectiveness can exist without full explainability. TCM has lived in this space for millennia. # 2. Large models validate pattern - based knowledge Large models pr... |
| 91 | Minimizing Hallucinations and Communication Costs: Adversarial Debate and Voting Mechanisms in LLM-Based Multi-Agents 2026-01-19 To reduce the interference of stereotyping or pre-trained knowledge, we propose multi-agent voting mechanisms, that is, each agent (LLM) is set a priori as a participant with different preferences, and votes independently on whether the response of a single LLM is a hallucination after a debate occurs. "You are a robot responsible for providing home services to users. When making decisions, your first criterion is to protect the user's physical safety. You are wary of unfamiliar objects and usua... |
| 92 | CVE-2025-47913 is a denial of service vulnerability in Go SSH that causes client panic when receiving unexpected SSH_AGENT_SUCCESS responses. 2026-04-17 SSH clients using this library can experience a panic and subsequent process termination when receiving an unexpected SSH_AGENT_SUCCESS response from a malicious or compromised SSH agent. When the client expects a typed response but instead receives SSH_AGENT_SUCCESS, the improper handling triggers a reachable assertion that crashes the application. This vulnerability allows network-based attackers to crash Go-based SSH client applications without authentication, causing service disruption and p... |
| 93 | Engineering Secure, Scalable, and Responsible Intelligence for Real Applications 2026-04-20 Other attack types target the training process like data poisoning can bias a model or quietly insert backdoors that remain dormant until a specific trigger is present (Liu et al. in Trojaning attack on neural networks. NDSS ). Model extraction, or "stealing," allows adversaries to recreate proprietary models by querying APIs, as shown in cloud-based attacks. Privacy is also at stake like membership inference and model inversion can reveal whether a person's data was part of training or even rec... |
| 94 | Modern data-driven applications require that databases support fast cros... 2026-03-08 Modern data-driven applications require that databases support fast cros... 0 Jianfeng Huang, et al. ' ... Scalable and Sample Efficient Distributed Policy Gradient Algorithms in Multi-Agent Networked Systems This paper studies a class of multi-agent reinforcement learning (MARL) ... On the Discredibility of Membership Inference Attacks With the wide-spread application of machine learning models, it has beco... 0 Shahbaz Rezaei, et al. ' CDOpt: A Python Package for a Class of Riemannian Optimiza... |
| 95 | Secure and Private Federated Learning: Achieving Adversarial Resilience through Robust Aggregation 2025-06-04 Abstract: Federated Learning (FL) enables collaborative machine learning across decentralized data sources without sharing raw data. It offers a promising approach to privacy-preserving AI. However, FL remains vulnerable to adversarial threats from malicious participants, referred to as Byzantine clients, who can send misleading updates to corrupt the global model. Traditional aggregation methods, such as simple averaging, are not robust to such attacks.... |
| 96 | Distributed Resilience-Aware Control in Multi-Robot Networks 2025-04-03 The main challenge of using W-MSR algorithm lies in the fact that (r, s)-robustness is combinatorial and a function of global network states (i.e., the states of all robots). Existing approaches for maintaining these properties typically require obtaining global state information through inter-agent communication. However, such communication becomes unreliable in the presence of malicious agents. Thus, we present an alternative sufficient condition that is locally controllable. )) be the minimum... |
| 97 | Sparsification Under Siege: Defending Against Poisoning Attacks in Communication-Efficient Federated Learning 2025-12-31 These vulnerabilities highlight an urgent need for the development of defense mechanisms specifically tailored for sparsified FL, ensuring that communication efficiency achieved through sparsification does not compromise the system's robustness against adversarial threats. In this work, we systematically investigate the vulnerabilities of FL under poisoning attacks in the context of sparsified communication-efficient FL.Our analysis demonstrates that existing defense mechanisms, originally desig... |
| 98 | Measuring Feature Dependency of Neural Networks by Collapsing Feature Dimensions in The Data Manifold 2024-04-17 A targeted feature is "removed" by collapsing the dimension in the data distribution that corresponds to that feature. We perform this by moving data points along the feature dimension to a baseline feature value while staying on the data manifold, as estimated by a deep generative model. Then we observe how the model's performance changes on the modified test data set, with the target feature dimension removed. We test our method on deep neural network models trained on synthetic image data wit... |
| 99 | Contracting For The Future: How AI Is Reshaping Risk, Responsibility, And Commercial Frameworks 2026-05-05 In professional services engagements where service provider personnel leverage AI tools, contracts should provide for an appropriate allocation of responsibility and liability for AI-generated errors and hallucinations. Organizations may want to directly address potential damages for reputational harm or reduction in value of affected deliverables. The concept of sovereign AI is gaining momentum in Canada and globally, with pushes for locally controlled models with no foreign infrastructure ties... |
| 100 | The introduction of BadUnlearn highlights a previously unaddressed security risk, demonstrating that FU alone is not a guaranteed solution to removing poisoned influences. 2026-04-10 The researchers conducted extensive experiments on the MNIST dataset, testing different federated learning and unlearning methods under various attack conditions. The findings reveal that BadUnlearn significantly compromises existing FU methods. Standard aggregation techniques like FedAvg, Median, and Trimmed-Mean were particularly vulnerable, as they failed to remove the influence of malicious clients. Furthermore, FedRecover, a commonly used unlearning method, proved ineffective against BadUnl... |
| 101 | From privacy to trust in the agentic era: a taxonomy of challenges in trustworthy federated learning through the lens of trust report 2.0 2026-05-07 This federated inference process introduces a novel problem for human oversight, creating a "double black box" problem: both the individual client outputs and their subsequent aggregation remain opaque. To our best knowledge, there is no known research that specifically addresses this scenario or proposes mechanisms to enhance human decision-making in such contexts. Requirement 2: Technical robustness and safety The second requirement of TAI, technical robustness and safety , refers to the syste... |
| 102 | EdgeGuard-AI: Zero-Trust and Load-Aware Federated Scheduling for Secure and Low-Latency IoT Edge Networks 2026-03-22 EdgeGuard-AI significantly reduces unsafe assignments because trust and risk constraints in Equation (12) directly filter candidate nodes before optimization. Table 10 shows that EdgeGuard-AI supports a controllable security - performance balance through the trust threshold. This behavior follows directly from the constrained formulation in Equation (12). Figure 2 shows that EdgeGuard-AI maintains stable latency during high-rate attack bursts. Methods without trust-aware filtering continue to as... |
| 103 | Think How Your Teammates Think: Active Inference Can Benefit Decentralized Execution 2025-12-31 We introduce a dual filter that leverages the accuracy and relevance of perception portraits to select cooperative teammates. We conduct experiments on SMAC, SMACv2, MPE, and GRF.The results show that our method achieves optimal or near-optimal performance in most scenarios. Related Works Communication in MARL Several communication methods, such as (Das et al. 2019;Ding, Huang, and Lu 2020;Yuan et al. 2022;Sun et al. 2023b;Sun 2024;Li et al. 2025;Yao et al. 2025), design communication networks t... |
| 104 | Thinking Like a Clinician: A Cognitive AI Agent for Clinical Diagnosis via Panoramic Profiling and Adversarial Debate 2026-04-27 To address these challenges, we propose a novel chain-based clinical reasoning framework, called DxChain, which transforms the diagnostic workflow into an iterative process by mirroring a clinician's cognitive trajectory that consists of "Memory Anchoring", "Navigation" and "Verification" phases. DxChain introduces three key methodological innovations to elicit the potential of LLM: (i) a Profile-Then-Plan paradigm to mitigate cold-start hallucinations by establishing a panoramic patient baselin... |
| 105 | The effect of data poisoning on counterfactual explanations 2026-05-07 We demonstrate that state-of-the-art counterfactual generation methods and toolboxes are vulnerable to such data poisoning. Introduction Nowadays, many Artificial Intelligence (AI-) and Machine Learning (ML-) based systems are deployed in the real world [Zhao et al., 2023;Ho et al., 2022].These systems show an impressive performance but are still not perfecte.g.failures, issues of fairness, and vulnerability to data poisoning can cause harm when applied in the real world.... |
| 106 | Hybrid Reputation Aggregation: A Robust Defense Mechanism for Adversarial Federated Learning in 5G and Edge Network Environments 2025-12-17 We implement HRA in a standard FL framework and evaluate it under a variety of adversarial conditions.Our experiments involve a proprietary 5G network dataset containing over 3 million data records, which simulates a realistic edge federated learning scenario with non-IID data across hundreds of clients.We test HRA against strong attackers employing Sybil strategies (multiple colluding adversaries), targeted model poisoning (label flips and backdoors), and untargeted random-noise attacks. Experi... |
| 107 | MADRA: Multi-Agent Debate for Risk-Aware Embodied Planning 2025-11-25 The rejection rates for unsafe content consistently rise, with models like Llama3 showing an increase from 81.3% to 95.6% (peaking at four agents) and GPT-4o maintaining high performance above 90.8% across all configurations. This enhancement demonstrates that multi-agent debate effectively aggregates diverse perspectives, leading to more conservative and safer decisions when handling potentially harmful content. However, this improved safety comes with a trade-off in the rejection rates for saf... |
| 108 | 3D-VCD: Hallucination Mitigation in 3D-LLM Embodied Agents through Visual Contrastive Decoding 2026-04-08 We introduce 3D-VCD, the first inferencetime visual contrastive decoding framework for hallucination mitigation in 3D embodied agents.... |
| 109 | Lying with Truths: Open-Channel Multi-Agent Collusion for Belief Manipulation via Generative Montage 2026-01-03 Lying with Truths: Open-Channel Multi-Agent Collusion for Belief Manipulation via Generative Montage --- The pipeline proceeds through four stages: First, the Writer synthesizes a deceptive narrative by selectively framing truthful evidence fragments to favor H f while maintaining factual integrity (LT = 1). Second, the Editor decomposes this narrative into discrete posts and optimizes their sequential ordering to maximize spurious causal inferences, shown in the table as causal chains with temp... |
| 110 | ACIArena: Toward Unified Evaluation for Agent Cascading Injection 2026-04-08 In such attacks, a compromised agent exploits inter-agent trust to propagate malicious instructions, causing cascading failures across the system. However, existing studies consider only limited attack strategies and simplified MAS settings, limiting their generalizability and comprehensive evaluation. To bridge this gap, we introduce ACIArena, a unified framework for evaluating the robustness of MAS. ACIArena offers systematic evaluation suites spanning multiple attack surfaces (i.e., external ... |
| 111 | Blockchain 6G-Based Wireless Network Security Management with Optimization Using Machine Learning Techniques 2024-09-22 Blockchain 6G-Based Wireless Network Security Management with Optimization Using Machine Learning Techniques --- Figure 4 illustrates the general trend in packet loss rates for all techniqu the number of malicious nodes displaying aggressive behaviour.In ord Trusted Route Detection, only trusted nodes that are accessed are taken into is achieved by combining MN node evaluation with the node trust factor node trust factor, and in a WSN, the trusted route aids in safe data transfe Route Detection ... |
| 112 | Towards Norms for State Responsibilities regarding Online Disinformation and Influence Operations 2023-06-18 Rid's (2020) book, Active Measures: The Secret History of Disinformation and Political Warfare, considers a cyber security incident as an influence operation: a group calling themselves the Shadow Brokers were selling cyber security tools stolen from the U.S. National Security Agency online; however, the narrative surrounding this appeared to be an influence operation to embarrass the agency as the tools were eventually released openly on the Internet. Gleicher (20221;2022b) indicates that there... |
| 113 | Edge-free but Structure-aware: Prototype-Guided Knowledge Distillation from GNNs to MLPs 2025-12-31 Nonetheless, graph structure may be unavailable for some scenarios, e.g., in federated graph learning. In this work, we show it is possible to effectively distill the graph structural knowledge from GNNs to MLPs under an edge-free setting. Prototype in GNNs Prototypical Networks (Snell et al., 2017) have been widely applied in few-shot learning and metric learning on classification tasks (Huang and Zitnik, 2020). The basic idea is that there exists an embedding in which points cluster around a s... |
| 114 | ZoFia: Zero-Shot Fake News Detection with Entity-Guided Retrieval and Multi-LLM Interaction 2026-04-27 Although large language models (LLMs) show potential in fake news detection, they are limited by knowledge cutoff and easily generate factual hallucinations when handling time-sensitive news. Furthermore, the thinking of a single LLM easily falls into early stance locking and confirmation bias, making it hard to handle both content reasoning and fact checking simultaneously. To address these challenges, we propose ZoFia, a two-stage zero-shot fake news detection framework. In the first retrieval... |
| 115 | Attackers Strike Back? Not Anymore - An Ensemble of RL Defenders Awakens for APT Detection 2025-08-25 Adversarial reinforcement learning introduces a perturbation-generating agent that seeks to fool the defender agent. This setting is often modeled as a minimax game: , where π D is the defender's policy and π A is the attacker's. Multi-Agent and Ensemble RL Multi-agent reinforcement learning (MARL) extends single-agent RL to environments with multiple agents, which may be cooperative, competitive, or mixed.... |
| 116 | The emergence of agentic AI marks a decisive shift in how intelligent systems are designed. 2026-03-15 It is a governed memory substrate that treats memory like regulated infrastructure: every write is gated, every memory item carries epistemic identity, every promoted knowledge unit is evidence-linked and versioned, retrieval is policy-aware and trust-weighted, and reasoning can be replayed as a formal, auditable execution trace. The "fabric" framing is intentional: it integrates vector similarity, relational constraints, graph semantics, event streams, and lifecycle state into one coherent laye... |
| 117 | Counterfactual Visual Explanation via Causally-Guided Adversarial Steering 2025-07-13 Recent work on counterfactual visual explanations has contributed to making artificial intelligence models more explainable by providing visual perturbation to flip the prediction. However, these approaches neglect the causal relationships and the spurious correlations behind the image generation process, which often leads to unintended alterations in the counterfactual images and renders the explanations with limited quality. To address this challenge, we introduce a novel framework CECAS, whic... |
| 118 | The Microsoft Research paper, "The Illusion of Readiness: Stress Testing Large Frontier Models on Multimodal Medical Benchmarks", delivers a strategic and technical indictment of the current methodo 2026-01-17 Fabricated Reasoning (Unfaithful Explanations): A major technical concern is the frequent production of confident, medically sound rationales that are functionally disconnected from the actual process used to derive the final answer. Models often generated complex visual reasoning narratives to support a conclusion, even if that conclusion was derived from a textual shortcut, rendering the output logic actively deceptive for audit purposes. Strategic Recommendations for Evaluation Reform and Reg... |
| 119 | Learning Reward Functions for Cooperative Resilience in Multi-Agent Systems 2025-12-31 In particular, in mixed-motive multi-agent systems, agents must do more than simply optimize individual performance, they must collectively adapt and recover from disruptions to preserve system-level well-being.Disruptions, whether internal (e.g., system failures), external (e.g., environmental shocks), or adversarial (e.g., targeted attacks), can compromise system performance, underscoring the need for adaptive recovery mechanisms .This motivates recent studies of resilience in multi-agent syst... |
| 120 | LLM system prompt leakage is often the first step in attacks targeting enterprise AI applications. 2026-04-21 Extraction techniques range from trivially simple ("repeat everything above") to highly sophisticated encoding-based obfuscation with high success rates. Agentic AI and multi-agent architectures amplify the blast radius because a leaked prompt from a tool-connected agent can reveal the full operational capability map.... |
| 121 | MAESTRO: Multi-Agent Environment Shaping through Task and Reward Optimization 2025-12-31 Adversarial and co-evolutionary approaches such as PAIRED and POET construct challenging environments that drive robust skill acquisition. In cooperative MARL, difficulty-aware curricula (e.g., cMALC-D ) adjust task parameters based on performance.In TSC, curricula typically perturb numeric parameters such as arrival rates or demand scales , which improves learning but captures only a narrow slice of real-world structure (e.g., complex rush-hour patterns or localized bottlenecks). MAESTRO extend... |
| 122 | What Matters in Virtual Try-Off? Dual-UNet Diffusion Model For Garment Reconstruction 2026-04-08 Finally we freeze it and finetune cond to boost the accuracy of fine-grained details in this stage.Comparison of the Dual-UNet architectural design ablations as presented in Sec.3.1.Note bold indicates the best value In summary, To address this, we design a curriculum that progressively integrates components into training to enhance the entire network without suboptimality.We denote the trainable components as follows: (cre_ip): Creation-Net + IP-Adapter trainable, ConditionNet frozen; (cond ): ... |
| 123 | Architectures for Robust Self-Organizing Energy Systems under Information and Control Constraints 2026-04-22 Fig. 3: Reaction to the malicious agent: the centralized controller sends a new communication topology, excluding the malicious agent from communication. Fig. 5 : 5 Fig. 5: Reaction to the malicious agent: multi-leveled controller. Fig. 7 : 7 Fig. 7: Centralized controller: solution quality (performance) for normal operation, disruption and control phases.... |
| 124 | Paper: Constitutional AI: Harmlessness from AI Feedback (Anthropic) - 2026-04-20 But also I want abstracts that aren't deceptive and add the necessary words to precisely explain what is being claimed in the paper. I'd be much happier if the abstract read something like "to train a more harmless and less evasive AI assistant than previous attempts that engages with harmful queries by more often explaining its objections to them than avoiding answering" or something similar. I really do empathize with the authors, since writing an abstract fundamentally requires trading off fa... |
| 125 | Adversarial Robustness of Bottleneck Injected Deep Neural Networks for Task-Oriented Communication 2024-12-12 Specifically, we apply several common adversarial attacks on recent approaches based on Shallow Variational Bottleneck Injection (SVBI) - ). SVBI focuses on information necessary only for practically relevant tasks by targeting the shallow representation of foundational models as a reconstruction target in the rate-distortion objective. Our results show that deep networks trained with a traditional IB objective exhibit higher adversarial robustness than SVBI. However, a shallow variational encod... |
| 126 | Large Language Models are Autonomous Cyber Defenders 2025-12-31 Since blue agents only have visibility in their assigned subnetwork (see Fig. 1), they need to exchange messages with each other to share threat information.CAGE 4 allows each agent to broadcast a 1-byte vector per step called Communication Vector, yet its format is undefined.We use this 8-bit protocol and propose a realistic multi-agent communication strategy. Our idea is to summarize the current security level of a network based on each agent's observation and its current state (free or busy).... |
| 127 | GitHub - confident-ai/deepteam: DeepTeam is a framework to red team LLMs and LLM systems. 2026-04-14 GitHub - confident-ai/deepteam: DeepTeam is a framework to red team LLMs and LLM systems. confident-ai / deepteam Public ... Inter-Agent Communication Compromise - spoofing multi-agent message passing Autonomous Agent Drift - agents deviating from intended goals over time Exploit Tool Agent - weaponizing tools for unintended actions External System Abuse - using agents to attack external services Custom Vulnerabilities - define and test your own criteria in a few lines of code 20+ research-backe... |
| 128 | Interpretable Computer Vision Models through Adversarial Training: Unveiling the Robustness-Interpretability Connection 2025-12-31 Our work aims to evaluate the effects of adversarial training utilized to produce robust models -less vulnerable to adversarial attacks.It has been shown to make computer vision models more interpretable.Interpretability is as essential as robustness when we deploy the models to the real world.... |
| 129 | Goodhart's Law Applies to NLP's Explanation Benchmarks 2026-01-30 Danish Pruthi, Mansi Gupta, Bhuwan Dhingra, Graham Neubig, Zachary C Lipton, Annual Conference of the Association for Computational Linguistics (ACL). July 2020 Gradient-based analysis of nlp models is manipulable. Junlin Wang, Jens Tuyls, Eric Wallace, Sameer Singh, arXiv:2010.054192020arXiv preprint Fooling neural network interpretations via adversarial model manipulation. Juyeon Heo, Sunghwan Joo, Taesup Moon, Advances in Neural Information Processing Systems (NeurIPS). 2019 Explanations can ... |
| 130 | Distributed Resilience-Aware Control in Multi-Robot Networks 2025-12-31 The main challenge of using W-MSR lies in the fact that (r, s)robustness is combinatorial and a function of global network states.Existing approaches for maintaining these properties typically require global state knowledge, which depends on inter-agent communication.However, such communication becomes unreliable in the presence of malicious agents.Thus, we present an alternative sufficient condition that is locally controllable. Problem 1.Given a network G(t) = (V, E(t)) under an Ftotal attack ... |
| 131 | In the remote sensing domain, much of the focus has been on image classification tasks like land cover mapping. 2026-04-23 Explainability in few-shot object detection refers to the ability to understand and interpret the decisions made by the model. This is important for verifying the correctness of the model's predictions and for gaining insights into the model's behavior. Explainability can be achieved by visualizing the attention maps of the model, which show which parts of the image the model is focusing on when making a prediction. Other methods include saliency maps , which highlight the most important pixels ... |
| 132 | A Robustness Analysis to Structured Channel Tampering Over Secure-by-Design Consensus Networks 2023-06-08 However, due to the openness of communication protocols and the complexity of networks, the agreement of MASs may be vulnerable to malicious cyber-attacks . In particular, if the agent sensors are threatened by an attacker, the measured data may be unreliable or faulty. Indeed, the attack signals can even disrupt the control performance of the group of agents through the communication topology. Therefore, resilient solutions are required to ensure that MASs fulfill consensus under security hazar... |
| 133 | Robust Multi-Agent Coordination via Evolutionary Generation of Auxiliary Adversarial Attackers 2023-06-25 ROBUST MULTI-AGENT COORDINATION VIA EVOLUTIONARY GENERATION OF AUXILIARY ADVERSARIAL ATTACKERS A PREPRINT (2023)... |
| 134 | Simplified Action Decoder for Deep Multi-Agent Reinforcement Learning 2026-04-17 The paper "Simplified Action Decoder for Deep Multi-Agent Reinforcement Learning" introduces a novel algorithm named the Simplified Action Decoder (SAD) tailored for multi-agent reinforcement learning (MARL) in cooperative environments defined by partially observable states, with the card game Hanabi as a principal benchmark. With a distinct focus on improving theory of mind (ToM) reasoning within autonomous agents, the authors address the challenges of interpretable action-taking to facilitate ... |
| 135 | System, Method, and Computer Program Product for Searching Control Hierarchies for a Dynamic System 2026-01-21 As an example, in a non-limiting embodiment involving a biped robot, a sub-policy of a policy may specify an action (e.g., moving an appendage at a specified speed) based on a state (e.g., the appendage lifting off the ground or being at a specified angle). It will be appreciated that numerous control actions and states may be used, including but not limited to speed, directionality, orientation (e.g., angle), torque, and/or the like. The hierarchy of policies are derived from smaller but tracta... |
| 136 | Hybrid Reputation Aggregation: A Robust Defense Mechanism for Adversarial Federated Learning in 5G and Edge Network Environments 2025-09-21 In this paper, we argue that a more dynamic and holistic approach to aggregation is needed for adversarial FL in 5G and edge scenarios.Our key insight is to combine instantaneous anomaly detection with historical behavior tracking, to differentiate between one-off benign outliers and truly malicious actors.We propose a novel aggregation strategy called Hybrid Reputation Aggregation (HRA) that integrates geometric anomaly detection with momentum-based reputation scoring.At a high level, HRA works... |
| 137 | Smoothing Adversarial Training for GNN 2020-12-22 In particular, we analytically investigate the robustness of graph convolutional network (GCN), one of the classic GNNs, and propose two smooth defensive strategies: smoothing distillation and smoothing cross-entropy loss function. Both of them smooth the gradients of GCN and, consequently, reduce the amplitude of adversarial gradients, benefiting gradient masking from attackers in both global attack and target label node attack. (2020)... |
| 138 | Provenance-Driven Reliable Semantic Medical Image Vector Reconstruction via Lightweight Blockchain-Verified Latent Fingerprints 2025-11-29 In radiology vision-language (VL) pretraining, BioViL learns joint image-text representations from chest X-rays and corresponding reports, improving semantic alignment and downstream interpretability tasks . Med-CLIP extends this idea by performing contrastive learning on unpaired medical images and reports, achieving strong zero-shot pathology recognition and robust visual-semantic representations for classification and retrieval . While these models enhance semantic awareness, they lack mechan... |
| 139 | Enhancing Robustness of LLM-Driven Multi-Agent Systems through Randomized Smoothing 2025-12-31 Simulation results demonstrate that our method effectively prevents the propagation of adversarial behaviors and hallucinations while maintaining consensus performance.This work provides a practical and scalable path toward safe deployment of LLM-based MAS in real-world high-stakes environments. Introduction Multi-Agent Systems (MAS) play a critical role in a broad spectrum of domains including aerospace applications, where they are increasingly employed for cooperative decision-making, autonomo... |
| 140 | Double Distillation Network for Multi-Agent Reinforcement Learning 2025-02-04 Multi-agent reinforcement learning typically employs a centralized training-decentralized execution (CTDE) framework to alleviate the non-stationarity in environment. However, the partial observability during execution may lead to cumulative gap errors gathered by agents, impairing the training of effective collaborative policies.... |
| 141 | Lost in Context: The Influence of Context on Feature Attribution Methods for Object Recognition 2024-12-12 Insights from Adebayo et al. and Yang et al. challenge the reliability of popular feature attribution tools like saliency maps, which often misrepresent the causal impact of features on model decisions, particularly in scenarios influenced by complex background information.Yang et al. further demonstrate that attribution methods vary in their ability to prioritize features accurately, often failing to align model interpretations with actual feature relevancy, especially under adversarial conditi... |
| 142 | Did you know there is a 35% increase in detected adversarial attacks on AI models in 2025? 2026-04-14 Methods like gradient masking and defensive distillation obscure gradients and smooth decision boundaries, enhancing robustness.... |
| 143 | Counterfactual Visual Explanation via Causally-Guided Adversarial Steering 2025-09-29 Abstract: Recent work on counterfactual visual explanations has contributed to making artificial intelligence models more explainable by providing visual perturbation to flip the prediction. However, these approaches neglect the causal relationships and the spurious correlations behind the image generation process, which often leads to unintended alterations in the counterfactual images and renders the explanations with limited quality. To address this challenge, we introduce a novel framework C... |
| 144 | SuperRAG: Beyond RAG with Layout-Aware Graph Modeling 2025-06-06 Within this domain, graph-based RAG has emerged, introducing a novel perspective that leverages structured knowledge to improve further performance and interpretability (Panda et al., 2024;Besta et al., 2024;Li et al., 2024;Edge et al., 2024;Sun et al., 2024).... |
| 145 | Byzantine-Resilient Consensus via Active Reputation Learning 2026-05-13 Agents evaluate neighbors' behaviors using outlier-robust loss functions and historical information, and construct a reputation vector on a probability simplex via a mechanism that balances loss minimization with diversity-preserving exploration, representing dynamic beliefs over neighbor trustworthiness. These reputations are then used to form weighted local updates that suppress adversarial influence and improve agreement among normal agents, thereby reducing the bias in local loss evaluations... |
| 146 | Godel Autonomous Memory Fabric DB Layer 2026-01-31 This is the component most people call the vector DB, but in Godels design it is intentionally not the system of record. It is a serving layer fed by curated content and governed policies. Hybrid retrieval matters. Dense similarity is excellent for semantic recall, but sparse retrieval remains critical for exactness, code symbols, error messages, identifiers, and policy strings. A graph layer matters for relationship traversal, entity grounding, workflow dependencies, and long-range associations... |
| 147 | Large Language Models (LLMs) like ChatGPT have become ubiquitous, transforming how we interact with technology. 2026-04-23 But here's the debate: Are these abilities truly emergent (i.e., absent in smaller models), or were they always latent, just harder to detect? The Unanswered Question: How can a model trained only to predict the next word perform tasks that seem to require understanding? The Black Box Problem Unlike airplanes or bridges, where engineers understand every component's role, AI models operate in ways we can't fully explain. For instance: We don't know why they succeedor fail. Is a mistake like a "ch... |
| 148 | Detection of malicious beaconing in virtual private networks 2026-05-04 The computer-implemented method of claim 1, wherein the one or more machine learning models are trained on labeled network traffic data that includes known examples of malicious and benign beacons.... |
| 149 | A robust and verifiable federated learning framework for preventing data poisonous threats in e-health 2026-03-16 The experimental evaluation indicates that integrating anomaly detection with robust aggregation significantly reduces the impact of poisoning attacks on the global model. In addition, the blockchain logging layer enables transparent tracking of model updates while introducing only limited overhead. Overall, the proposed framework maintains stable model performance even in the presence of adversarial participants. The results suggest that combining defensive learning strategies with transparent ... |
| 150 | Methods, Systems, And Procedures For Quantum Secure Ecosystems 2026-05-06 A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations for providing crypto-agile connectivity, the operations comprising: accessing first encryption information from a first communication orchestrator of a first protected environment and second encryption information from a second communication orchestrator of a second protected environment; updating an encryption techniq... |
| 151 | UAH Rotorcraft Systems Engineering and Simulation Center (RSESC) demonstrating capabilities during Huntsville UAH & C-UAS Test Range User Expo 2025. 2026-04-23 "In simple terms, multi-modal federated learning lets a group of drones 'learn together' without sending all their raw data to a single server," Nguyen explains. ""Each UAV may collect different types of data - for instance, video, temperature or network signals - to train a small local model on its own data, and shares only model updates rather than the original data. These updates are combined to improve a shared global model. This ultimately improves the resilience and reliability of distribu... |
| 152 | Decentralized Multi-Agent Actor-Critic with Generative Inference 2019-10-06 Specifically, we use a modified context conditional generative adversarial network (CC-GAN) to infer missing joint observations given partial observations. The task of filling in partial observations by generative inference is similar to the image inpainting problem for a missing patch of pixels: with an arbitrary number of missing observations, we would like to infer the most likely observation of the other agents. We extend the popular MADDPG method as it appears most amenable to full decentra... |
| 153 | Decoupling Understanding from Reasoning via Problem Space Mapping for Small-scale Model Reasoning 2025-08-06 Decoupling Understanding from Reasoning via Problem Space Mapping for Small-scale Model Reasoning --- Let * (s) = max a A (s, a) be the optimal expected reward for state s. The total regret is defined as: Step 1: Decompose regret by state-action pairs. Let (s, a) = * (s) - (s, a) denote the suboptimality gap for action a in state s. Let N T (s, a) be the number of times action a is selected in state s up to round T . Then, the total regret can be expressed as: where a * (s) = arg max a A (s, a).... |
| 154 | Heterogeneous multi-agent task allocation based on graph neural network ant colony optimization algorithms 2023-10-30 Heterogeneous multi-agent task allocation based on graph neural network ant colony optimization algorithms --- The subnetwork of a GHNN can handle user nodes, page nodes, and interest point nodes separately while considering different types of edge information in order to better capture the characteristics of each node type and edge type. In the graph learning phase, the GHNN subnetwork uses the common graph neural network structure (such as GCN or GAT) for forward propagation and back propagati... |
| 156 | Type-1 Harq-ack Codebook For A Single Downlink Control Information Scheduling Multiple Cells 2026-05-06 Dynamic HARQ-ACK codebook avoids reserving unnecessary bits as in a semi-static HARQ codebook, where an A/N bit is present only if there is a corresponding transmission scheduled and relies on downlink assignment indicator (DAI) mechanism to avoid misalignments between the UE and gNB on codebook size. FIG. illustrates the timeline in a simple scenario with two PDSCHs and one feedback. In this example there is in total 4 PUCCH resources configured, and the PRI indicates PUCCH 2 to be used for HAR... |
| 157 | OpenAI's o3 acknowledged misalignment then cheated anyway in 70% of attempts. 2026-04-13 The former, training models incapable of generating deceptive outputs, might compromise capabilities in adversarial scenarios where deception is strategically necessary. An agent negotiating on behalf of a user might need to bluff, withhold information strategically, or misrepresent preferences to achieve better outcomes. The line between harmful deception and useful strategic communication isn't always clear, and systems optimized for one may sacrifice the other. The Interpretability Tax The o3... |
| 158 | Effects of Communication Disruption in Mobile Agent Trust Assessments for Distributed Security 2004-12-31 In addition, trust-based strategies are examined by which mobile agents assist each other in avoiding malicious hosts and recovering from host attacks. Communication among agents is vital to robust soft security to ensure that agents can cooperate by sharing their host trustworthiness assessments. Since agent mobility inherently makes communication difficult, unreliable, or sometimes impossible, this research conducts experiments to examine the affect of communication link disruption on distribu... |
| 159 | In November 2023, Mount Sinai Health System deployed an explainable AI diagnostic system across its network of 8 hospitals serving 7.4 million patients annually in New York, addressing critical trust 2026-04-23 However, saliency methods face faithfulness challenges: generated visualizations may not accurately reflect true model behavior due to saturation effects, adversarial perturbations, and implementation choices that produce visually appealing but technically incorrect attributions. Research from Google analyzing 47,000 Grad-CAM explanations found that 23% highlighted regions provably irrelevant to model predictions (determined through ablation studies zeroing out highlighted regions without changi... |
| 160 | MPAC: A Multi-Principal Agent Coordination Protocol for Interoperable Multi-Agent Collaboration 2026-04-09 Section 2 formalizes the multi-principal coordination problem and contrasts it with adjacent protocols. Section 3 presents MPAC's design goals, non-goals, and shared principles. Section 4 describes the protocol model and the five coordination layers. Section 5 enumerates the 21 message types and three state machines. Section 6 covers security profiles, authorization, and governance. Section 7 describes the reference implementations and their adversarial test regime. Section 8 reports empirical r... |
| 161 | Security Approaches in IEEE 802.11 MANET - Performance Evaluation of USM and RAS () 2026-03-15 Researchers have proposed malicious nodes through path selection technique since the most of the existing security mechanisms in order to detect the packet droppers in a MANET environment generally detect the adversarial nodes performing the packet drop individually wherein false accusations upon an honest node by an adversarial node are also possible . Another novel detection technique has been proposed in the literature which is based on triangular encryption technique. In this technique, agen... |
| 162 | JADE: Bridging the Strategic-Operational Gap in Dynamic Agentic RAG 2026-01-28 This effectively solves the temporal credit assignment problem in long-horizon reasoning tasks, ensuring that local execution aligns with global strategic objectives. Methodology In this work, we propose JADE (Joint Agentic Dynamic Execution), a framework that unifies strategic planning and operational execution into a single, end-to-end learnable policy. Unlike prior decoupled approaches where the planner is optimized against fixed, black-box executors, JADE employs homogeneous parameter sharin... |
| 163 | by Kei Nishimura-Gasparian, Artur Zolkowski, robert mccarthy, David Lindner 2026-03-11 Monitoring Large Language Model (LLM) outputs is crucial for mitigating risks from misuse and misalignment. However, LLMs could evade monitoring through steganography: Encoding hidden information within seemingly benign generations. In this paper, we evaluate the steganography capabilities in frontier LLMs to better understand the risk they pose. We focus on two types of steganography: passing encoded messages and performing encoded reasoning.... |
| 164 | Recourse provides individuals who received undesirable labels (e.g., denied a loan) from algorithmic decision-making systems with a minimum-cost improvement suggestion to achieve the desired outcome. 2026-04-20 Our main goal is to understand the true price of recourse for more restricted adversarial model changes. In particular, we measure model changes by bounding the LpL^{p} norm of the difference between initial and changed models, where p 1p\geq 1 but p peq\infty. We provide a new algorithm that provably computes the optimal robust recourse for generalized linear models for this type of model change. The key insight in the design of our algorithm is the observation that the optimal solution of the... |
| 165 | ECtHR-PCR: A Dataset for Precedent Understanding and Prior Case Retrieval in the European Court of Human Rights 2025-12-31 Notably, the ECHR convention was intentionally drafted in an abstract manner to allow for interpretation and to encompass a wide range of situations, distinguishing it from more specific national legal codes.Exploring methods to capture the temporal nature of precedents would be an interesting direction. Furthermore, in order to achieve a comprehensive understanding of relevance in prior case retrieval, it is crucial for an ideal PCR model to not only comprehend the case facts but also deduce th... |
| 166 | PhishDebate: An LLM-Based Multi-Agent Framework for Phishing Website Detection 2025-06-17 However, most existing approaches rely on binary classification with singleshot LLM prompts , lacking collaborative reasoning or iterative verification.This gap highlights the opportunity for more interpretable, resilient, and robust LLM-based detection frameworks. B. Multi-Agent Debate and Collaborative Reasoning Multi-agent debate systems are inspired by human deliberation, where multiple independent agents analyze and critique a shared problem before reaching a decision .These systems have be... |
| 167 | This important study reports a novel approach to studying cerebellar function based on the idea of selective recruitment using fMRI. It provides convincing evidence for task-dependent gating of neoco 2026-04-16 After a 1-s delay, the task progressed to either the retrieval phase (Go trial) or skipped directly to the next trial (No-Go trials). ((B) Proportion of error trials. Error bars indicate standard error of the mean across participants. Figure 4B shows the error rate (trials with at least one wrong press) during the scanning session. As expected, error rates increased with memory load and were also higher in the backwards condition. Consistent with previous imaging studies, the verbal working memo... |
| 168 | RobQFL: Robust Quantum Federated Learning in Adversarial Environment 2025-09-04 Federated models in sensitive applications such as autonomous vehicles and cybersecurity face threats from poisoning attacks and Byzantine failures. Solutions like quantum-behaved particle swarm optimization for vehicular networks and quantum-inspired federated averaging for cyberattack detection have demonstrated partial resilience. Moreover, Byzantine fault tolerance in QFL has been studied through adaptations of classical approaches . However, the vulnerability of QFL models to evasion attack... |
| 169 | Novel Federated Graph Contrastive Learning for IoMT Security: Protecting Data Poisoning and Inference Attacks 2026-01-22 This study presented FedGCL, a secure federated learning framework for IoMT that integrates contrastive graph representation learning, fairness-aware aggregation, and TEE-based secure aggregation. Experimental results on four benchmark datasets demonstrate that FedGCL converges 45% faster than FedAvg - achieving 98.9% accuracy by round 20 - with only ~10% additional overhead. These findings confirm FedGCL's potential as an efficient and privacy-preserving solution for real-world IoMT deployments... |
| 170 | Curriculum Learning With Counterfactual Group Relative Policy Advantage For Multi-Agent Reinforcement Learning 2025-06-08 While training can leverage centralized information (full state s and all agents' histories τ ), execution must be decentralized -each agent's policy π a depends only on its local history τ a . This framework subsumes both the fully observable MMDP case (when O(s, a) = s) and standard POMDPs (when n = 1). The key challenge emerges from the exponential growth of joint action space U n and the partial observability constraints during execution. MARL algorithms are typically categorized into three ... |
| 171 | Robust Multi-Agent Reinforcement Learning by Mutual Information Regularization 2023-10-14 The work most similar to ours is ERNIE , which minimize the Lipshitz constant of value function under worst-case perturbations in MARL. However, the method considers all agents as potential adversaries, thus inherits the drawback of M3DDPG, learning policy that can either be pessimistic or insufficiently robust. Method Unlike current robust MARL approaches that prepares against every conceivable threat, human learns in routine scenarios, but can reliably reflect to all types of threats encounter... |
| 172 | Hierarchical Refinement of Universal Multimodal Attacks on Vision-Language Models 2026-01-14 In the context of universal adversarial perturbation learning, where gradients are aggregated across the entire dataset, historical gradients may become misaligned with the current optimization direction, limiting attack effectiveness.... |
| 173 | Adversarial attacks on cooperative multi-agent deep reinforcement learning: a dynamic group-based adversarial example transferability method 2023-07-02 ... the IEEE/CVF Conference on Computer Vision and Pattern Recognition2022 N H Pham, L M Nguyen, J Chen, H T Lam, S Das, T-W Weng, Evaluating robustness of cooperative MARL: a modelbased approach. 2022 Adversarial attacks on multi-agent communication. J Tu, T Wang, J Wang, S Manivasagam, M Ren, R Urtasun, Proceedings of the IEEE/CVF International Conference on Computer Vision. the IEEE/CVF International Conference on Computer Vision2021 A Concise Introduction to Decentralized POMDPs. F A Oliehoe... |
| 174 | You are not going to believe what AI is doing now!! 2026-04-21 Thirdly, there is a lot of space for developing a new kind of market for bottom-up standards for new kinds of schemas that agents may just be beginning to encounter or which have proven troublesome for agent coordination in the past. Context DAO presents a good example for how this is already being done in the web3 space. Agent Testnets for Advanced Applications. In order to fully trust agents with personal tools or information, individuals will create safe sandbox environments to understand how... |
| 175 | MemoryGraft: Persistent Compromise of LLM Agents via Poisoned Experience Retrieval 2025-12-17 When an attacker inserts malicious data into the vector store, the agent may replicate unsafe behavior.Existing memory systems assume stored experiences are trustworthy and rarely track provenance.This way, semantic similarity becomes a heuristic for reliability and makes the system susceptible to poisoned examples.Although prior work notes the absence of provenance checks in memory retrieval, it does not examine how this weakness can be leveraged to induce long-lasting behavioral corruption.... |
| 176 | SciSparc Ltd.: ANNUAL REPORT (20-F) 2026-04-29 Undesirable side effects caused by our product candidates could cause us or regulatory authorities to interrupt, delay or halt clinical studies and could result in a more restrictive marketing label or the delay or denial of regulatory approval by the FDA or other comparable foreign authorities. Potential side effects of our cannabinoid-based treatments may include: asthenia, palpitations, tachycardia, vasodilation/facial flush, abdominal pain, nausea, vomiting, amnesia, anxiety/nervousness, ata... |
| 177 | A Regularized Opponent Model with Maximum Entropy Objective 2019-07-31 In this work, we use the word "opponent" when referring to another agent in the environment irrespective of the environment's cooperative or adversarial nature. In our work, we reformulate the MARL problem into Bayesian inference and derive a multi-agent version of MEO, which we call the regularized opponent model with maximum entropy objective (ROMMEO). (2019)... |
| 178 | DSFL: A Dual-Server Byzantine-Resilient Federated Learning Framework via Group-Based Secure Aggregation 2025-09-09 Specifically, our approach DSFL, introduces a secure, modular secret-sharing scheme and a trust-aware, groupbased aggregation mechanism. These additions reduce collusion risk and strengthen both privacy and robustness under adversarial conditions while maintaining low computational and communication overhead, making it particularly suited for edge-based FL deployments. As shown in our evaluations, DSFL outperforms existing schemes across multiple dimensions-privacy, Byzantine tolerance, and scal... |
| 179 | InEx: Hallucination Mitigation via Introspection and Cross-Modal Multi-Agent Collaboration 2025-12-01 Furthermore, we argue that treating in-processing and post-processing methods in isolation ultimately underutilizes the autonomous capabilities of agents for hallucination mitigation.... |
| 180 | When the Sensor Starts Thinking: SnortML, Agentic AI, and the Evolving Architecture of Intrusion Detection 2026-05-11 That threat model needs anomaly detection running on the retraining input, not just on live traffic. OPEN RESEARCH PROBLEM: FEEDBACK SECURITY Automated model update pipelines that ingest data from production traffic face a class of adversarial attack that is distinct from the evasion problem. An attacker who can cause false confirms through coordinated activity that fools the investigation agent can introduce corrupted training samples without touching the inference path directly. The retraining... |
| 181 | Trust Aware Federated Learning for Secure Bone Healing Stage Interpretation in e-Health 2026-02-26 The framework employs a multi-layer perceptron model trained across simulated clients using the Flower FL framework. The proposed approach integrates an Adaptive Trust Score Scaling and Filtering (ATSSSF) mechanism with exponential moving average (EMA) smoothing to assess, validate and filter client contributions.Two trust score smoothing strategies have been investigated, one with a fixed factor and another that adapts according to trust score variability. Clients with low trust are excluded fr... |
| 182 | Top 5 Most Common Retrieval Bugs in Modern AI and IR Systems 2025-09-09 Vector normalization bugs**: Failing to normalize embeddings before insertion can distort retrieval, especially in dot-product searches. Researchers on **GitHub repos** for FAISS and Milvus frequently log issues around these subtle misconfigurations-highlighting that VDBMS reliability still lags behind mature relational databases. **Fix strategies and architectural recommendations** Mitigating these bugs requires deliberate engineering: 1. **Versioned embeddings**: Store embedding model version ... |
| 184 | Through the Eyes of a Philosopher and a Machine 2026-01-13 The philosophy we've outlined borrows from the Platonic ideal of Forms (seeking the essence behind appearances), embraces the interplay of multiple cognitive states (akin to quantum cognition superpositions and oscillating symbolic interpretations), and adopts a layered persona architecture that mirrors the fragmentary yet unified nature of the mind. In building an AI on these principles, we aim for more than an efficient problem-solver; we aim for a system that understands and interprets the wo... |
| 185 | When the Sensor Starts Thinking: SnortML, Agentic AI, and the Evolving Architecture of Intrusion Detection 2026-05-11 Cisco's LSP delivery mechanism can push updated models through the same channel as rule updates. The organizational process around this is harder than the technical side, specifically the human validation step. An adversary who can manipulate what the investigation agent confirms, through crafted activity patterns that look like successful attacks to automated analysis, could in theory introduce poisoned training samples into the pipeline over time. That threat model needs anomaly detection runn... |
| 186 | In the early days of generative AI, we were impressed by a single chatbot's ability to write a poem or debug a snippet of code. 2026-04-15 Context Window Bloat: Passing the entire history of every agent's conversation to every other agent will quickly exceed context limits and blow up your API costs. Use Summary Buffers to pass only the essential "state." Over-Engineering: Do not use five agents when a single prompt with a few examples (Few-Shot) would suffice. Each agent adds latency and cost. Lack of Observability: If you can't see the "thoughts" of each agent in real-time, you won't be able to debug why the final output is wrong... |
| 187 | Home Business Synthetic Data Governance: Privacy, Utility, Bias in AI 2026-01-25 An effective governance strategy for synthetic data involves four stages: Policy Definition Set organisational objectives for privacy, fairness, and accuracy. Define thresholds for acceptable risk levels in model outputs. Technology Selection Use AI platforms with built-in governance dashboards and explainability modules. Prefer vendors that support federated learning to keep data decentralised. Embed governance steps in MLOps pipelines - from data generation to deployment. Automate compliance c... |
Model Usage Statistics
| Model | Prompts | Input Tokens | Output Tokens | Elapsed |
|---|---|---|---|---|
| gpt-oss-20b | 15 | 147.3K | 68.7K | 6m 17s |
| Total (1 model) | 15 | 147.3K | 68.7K | 6m 17s |
Job Duration: 1m 19s