When Actions Disappear: Adversarial Action Removal in Self-Play Reinforcement Learning
arXiv:2605.16312v1 Announce Type: new
Abstract: We study adversarial action masking in self-play reinforcement learning: an attacker selectively removes legal actions from a victim's action set. Unlike observation or action perturbations, removal eliminates decision options before the agent acts. A...
AgentWall: A Runtime Safety Layer for Local AI Agents
arXiv:2605.16265v1 Announce Type: new
Abstract: The safety of autonomous AI agents is increasingly recognized as a critical open problem. As agents transition from passive text generators to active actors capable of executing shell commands, modifying files, calling APIs, and browsing the web, the ...
ANNEAL: Adapting LLM Agents via Governed Symbolic Patch Learning
arXiv:2605.16309v1 Announce Type: new
Abstract: LLM-based agents can recover from individual execution errors, yet they repeatedly fail on the same fault when the underlying process knowledge--operator schemas, preconditions, and constraints--remains unrepaired. Existing self-evolving approaches ad...
From Prompts to Protocols: An AI Agent for Laboratory Automation
arXiv:2605.16552v1 Announce Type: new
Abstract: Automating science laboratories enables faster, safer, more accurate, and more reproducible execution of protocols, accelerating the discovery and testing of new materials, drugs, and more. However, setting up and running autonomous labs requires coor...
Skim: Speculative Execution for Fast and Efficient Web Agents
arXiv:2605.16565v2 Announce Type: new
Abstract: Skim is a speculative execution framework for web agents that exploits the predictable structure of purpose-built websites. Today's web-agent expense is not intrinsic to the tasks but a property of how agents are composed: frontier-model inference, br...
Scalable Uncertainty Reasoning in Knowledge Graphs
arXiv:2605.16568v1 Announce Type: new
Abstract: Knowledge Graphs are pivotal for semantic data integration. The real-world data they model is often inherently uncertain. Within knowledge graphs, uncertainty manifests in three distinct levels: imprecise attribute values, probabilistic triple existen...
EpiCache: Episodic KV Cache Management for Long-Term Conversation on Resource-Constrained Environments
Modern large language models (LLMs) extend context lengths to millions of tokens, enabling coherent, personalized responses grounded in long conversational history. However, the Key-Value (KV) cache grows linearly with the extended dialogue history, causing the model’s memory footprint to quickly ex...
AgentStop: Terminating Local AI Agents Early to Save Energy in Consumer Devices
arXiv:2605.15206v1 Announce Type: new
Abstract: Autonomous agents powered by large language models (LLMs) are increasingly used to automate complex, multi-step tasks such as coding or web-based question answering. While remote, cloud-based agents offer scalability and ease of deployment, they raise...
TeamTR: Trust-Region Fine-Tuning for Multi-Agent LLM Coordination
arXiv:2605.15207v1 Announce Type: new
Abstract: Multi-agent LLM systems have shown promise for complex reasoning, yet recent evaluations reveal they often underperform single-model baselines. We identify a structural failure mode in sequential fine-tuning of shared-context teams: updating one agent...
Quantization Undoes Alignment: Bias Emergence in Compressed LLMs Across Models and Precision Levels
arXiv:2605.15208v1 Announce Type: new
Abstract: Large Language Models are routinely compressed via post-training quantization to reduce inference costs and memory footprint for cloud and edge deployment, yet the impact of this compression on model quality remains poorly understood. Existing studies...
Mask-Morph Graph U-Net: A Generalisable Mesh-Based Surrogate for Crashworthiness Field Prediction under Large Geometric Variation
arXiv:2605.15231v1 Announce Type: new
Abstract: Nonlinear finite element crash simulations are accurate but computationally expensive, limiting their use in iterative design optimisation. Machine-learning surrogate models based on graph neural networks (GNNs) offer a faster alternative. Message-pas...
MuteBench: Modality Unavailability Tolerance Evaluation for Incomplete Multimodal Fusion
arXiv:2605.15235v1 Announce Type: new
Abstract: Multimodal physiological data powers clinical AI systems from intensive care units to wearable devices, but sensors routinely fail in practice. Two failure modes are common: modality missing, where an entire channel is absent, and within-modality miss...
DeepSlide: From Artifacts to Presentation Delivery
arXiv:2605.15202v1 Announce Type: new
Abstract: Presentations are a primary medium for scholarly communication, yet most AI slide generators optimize the artifact (a visually plausible deck) while under-optimizing the delivery process (pacing, narrative, and presentation preparation). We present De...
SDOF: Taming the Alignment Tax in Multi-Agent Orchestration with State-Constrained Dispatch
arXiv:2605.15204v1 Announce Type: new
Abstract: Multi-agent orchestration frameworks such as LangChain, LangGraph, and CrewAI route tasks through graph-based pipelines but do not enforce the stage constraints that govern real business processes. We present SDOF, a framework that treats multi-agent ...
Does Theory of Mind Improvement Really Benefit Human-AI Interactions? Empirical Findings from Interactive Evaluations
arXiv:2605.15205v1 Announce Type: new
Abstract: Improving the Theory of Mind (ToM) capability of Large Language Models (LLMs) is crucial for effective social interactions between these AI models and humans. However, the existing benchmarks often measure ToM capability improvement through story-read...
Fair outputs, Biased Internals: Causal Potency and Asymmetry of Latent Bias in LLMs for High-Stakes Decisions
arXiv:2605.15217v1 Announce Type: new
Abstract: Instruction-tuned language models exhibit behavioural fairness in high-stakes decisions while retaining biased associations in their internal representations. However, whether these suppressed representations can affect model outputs - and whether suc...
GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration
arXiv:2605.13848v1 Announce Type: new
Abstract: Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routing, infinite loops, and non-reproducible execution. We introduce GraphBit, an engine-orchestrated f...
Mixed Integer Goal Programming for Personalized Meal Optimization with User-Defined Serving Granularity
arXiv:2605.13849v1 Announce Type: new
Abstract: Determining what to eat to satisfy nutritional requirements is one of the oldest optimization problems in operations research, yet existing formulations have two persistent limitations: continuous variables produce impractical fractional servings (1.7...
A Two-Dimensional Framework for AI Agent Design Patterns: Cognitive Function and Execution Topology
arXiv:2605.13850v1 Announce Type: new
Abstract: Existing frameworks for LLM-based agent architectures describe systems from a single perspective: industry guides (Anthropic, Google, LangChain) focus on execution topology -- how data flows -- while cognitive science surveys focus on cognitive functi...
Invisible Orchestrators Suppress Protective Behavior and Dissociate Power-Holders: Safety Risks in Multi-Agent LLM Systems
arXiv:2605.13851v1 Announce Type: new
Abstract: Multi-agent orchestration -- in which a hidden coordinator manages specialized worker agents -- is becoming the default architecture for enterprise AI deployment, yet the safety implications of orchestrator invisibility have never been empirically tes...
arXiv:2605.13880v1 Announce Type: new
Abstract: Agent memory is typically constructed either offline from curated demonstrations or online from post-deployment interactions. However, regardless of how it is built, an agent faces a cold-start gap when first introduced to a new environment without an...
Further Notes on Our Recent Research on AI Delegation and Long-Horizon Reliability
Our recent paper, “LLMs Corrupt Your Documents When You Delegate”, has generated discussion about the reliability of AI systems in delegated workflows. We appreciate the interest in this work and want to clarify several important points about what the paper does—and does not—claim. The research aims...
Mechanistic Interpretability of EEG Foundation Models via Sparse Autoencoders
arXiv:2605.13930v1 Announce Type: new
Abstract: EEG foundation models achieve state-of-the-art clinical performance, yet the internal computations driving their predictions remain opaque: a barrier to clinical trust. We apply TopK Sparse Autoencoders (SAEs) across three architecturally distinct EEG...