AI news archive.
19,864 articles, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchECG-WM: A Physiology-Informed ECG World Model for Clinical Intervention Simulation
arXiv:2605.17580v2 Announce Type: replace Abstract: Electrocardiogram (ECG)-based models have achieved strong performance in diagnostic tasks, yet they remain limited in modeling how cardiac dynamics evolve under externa…
- arXiv cs.AIagentsAgentic Trading: When LLM Agents Meet Financial Markets
arXiv:2605.19337v2 Announce Type: replace Abstract: Large Language Models (LLMs) combined with autonomous agent architectures are shifting quantitative finance from isolated predictive modeling toward closed-loop trading…
- arXiv cs.AIagentsTowards Direct Evaluation of Harness Optimizers via Priority Ranking
arXiv:2605.22505v3 Announce Type: replace Abstract: Harness optimization enables automated agent creation by having an optimizer agent iteratively update the harness of target agents. Despite its success, current studies…
- arXiv cs.AIresearchDr-CiK: A Benchmark testing Deep Research for Context-Aided Forecasting
arXiv:2605.27904v2 Announce Type: replace Abstract: Time series forecasting in real-world settings often depends not only on historical observations, but also on external context that must be actively discovered from noi…
- arXiv cs.AIagentsPlant, Persist, Trigger: Sleeper Attack on Large Language Model Agents
arXiv:2605.28201v2 Announce Type: replace Abstract: Large Language Model (LLM) agents remain vulnerable to safety threats from the external environment, where attackers inject adversarial content into external observatio…
- arXiv cs.AIresearchMEMENTO: Leveraging Web as a Learning Signal for Low-Data Domains
arXiv:2605.29795v2 Announce Type: replace Abstract: Real-world tasks often lack large labeled datasets, motivating extensive work on learning in low-data regimes. Existing approaches such as few-shot prompting, instructi…
- arXiv cs.AIagentsCharacterizing How Complex Agentic AI Systems Handle General Tasks: A Trace-Based Simulation Study
arXiv:2606.01725v2 Announce Type: replace Abstract: Agentic AI completes tasks through iterative planning, tool use, and reasoning based on observed outcomes. Despite its popularity, its system-level behavior remains poo…
- arXiv cs.AIresearchBeyond Objective Equivalence: Constraint Injection for LLM-Based Optimization Modeling on Vehicle Routing Problems
arXiv:2606.04816v2 Announce Type: replace Abstract: Large language models (LLMs) can generate executable solver code from natural-language descriptions of optimization problems. However, existing verification signals foc…
- arXiv cs.AIresearchScaling Participation in Modular AI Systems
arXiv:2606.07812v2 Announce Type: replace Abstract: Humanity is a mosaic of multifaceted talents and needs, and any truly intelligent AI must reflect that richness. Yet the LLMs used by all are built by the few -- a cent…
- arXiv cs.AIresearchSafety is Contextual, LLM-Judges Are Not: Navigating the Rigid Priors of Evaluators
arXiv:2606.07874v2 Announce Type: replace Abstract: LLMs-as-judges are the primary way to evaluate safety at scale. Despite their importance, LLM-judges themselves are rarely evaluated beyond human agreement in simple, s…
- arXiv cs.AIresearchCIAware-Bench: Benchmarking Control Intervention Awareness Across Frontier LLMs
arXiv:2606.11063v2 Announce Type: replace Abstract: AI control protocols oversee untrusted models by monitoring their actions and modifying potentially unsafe steps, often using a trusted model. This partially tampers wi…
- arXiv cs.AIagentsSelfEvoSkill: Paired Execution Audits for Skill Revision without External Outcome Supervision
arXiv:2606.14239v2 Announce Type: replace Abstract: Agent skills provide LLM agents with reusable procedures, but a given skill may be incomplete, ineffective, or misleading for the current task. Existing methods for rev…
- arXiv cs.AIagentsRTSGameBench: An RTS Benchmark for Strategic Reasoning by Vision-Language Models
arXiv:2606.18950v3 Announce Type: replace Abstract: Modern Vision-Language Models (VLMs) often struggle with strategic reasoning, i.e., anticipating and influencing other agents' actions, under uncertainty in competitive…
- arXiv cs.AIagentsAgent MechSuits: Mechanistic Subspace Safety Steering for Multi-Turn CLI Agents
arXiv:2606.22673v2 Announce Type: replace Abstract: Command-Line Interface (CLI) agents based on large language models (LLMs) demonstrate remarkable autonomous capabilities, but they also introduce significant safety and…
- arXiv cs.AIresearchDART: Draft-Agreement Routing for Training-Free Adaptive Thinking Budgets in Hybrid Reasoning Models
arXiv:2606.23181v4 Announce Type: replace Abstract: Hybrid reasoning models can answer directly or spend extra tokens on extended thinking. A practical router should choose between these modes for each query, so easy pro…
- arXiv cs.AIresearchTIPS: Topological Ill-Posedness Probing and Steering in Large Language Models
arXiv:2606.23590v2 Announce Type: replace Abstract: Ill-posed questions, including those involving ambiguity, under-specification, or conflicting statements, may admit no valid answer or multiple plausible answers, posin…
- arXiv cs.AIresearchMORPH: Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimization
arXiv:2606.26899v2 Announce Type: replace Abstract: Embedding-based retrieval typically returns highest-scoring items, but many production scenarios require items that satisfy a target attribute while preserving a fine-g…
- arXiv cs.AIresearchLook-Before-Move: Narrative-Grounded World Visual Attention in Dynamic 3D Story Worlds
arXiv:2606.26964v3 Announce Type: replace Abstract: As embodied AI and world models increasingly operate in dynamic 3D environments, visual perception must move beyond passively interpreting given observations toward act…
- arXiv cs.AIresearchBeyond Compilation: Evaluating Faithful Natural-Language-to-Lean Statement Formalization
arXiv:2606.31002v4 Announce Type: replace Abstract: Lean verifies that a generated declaration is well typed, but cannot check whether it states what the user intended. We measure how far compilation overstates faithfuln…
- arXiv cs.AIagentsMastermind: Strategy-grounded Learning for Repository-Scale Vulnerability Reproduction
arXiv:2607.01764v2 Announce Type: replace Abstract: Repository-level vulnerability reproduction is a demanding software engineering (SE) task: an agent must inspect a codebase, infer the input grammar that reaches a vuln…
- arXiv cs.AIresearchSubliminal Clocks: Latent Time Modelling in Diffusion Language Models
arXiv:2607.01774v3 Announce Type: replace Abstract: Diffusion Language Models (DLMs) have recently emerged as a promising alternative to autoregressive models. Unlike standard diffusion-based approaches, DLMs are not exp…
- arXiv cs.AIagentsAgenticPD: A Stage-Aware Agentic Framework for Closed-Loop Physical Design Optimization
arXiv:2607.04758v3 Announce Type: replace Abstract: Physical design quality-of-results (QoR) optimization is hard and expensive. Choices made at one stage can help or hurt later stages. Each evaluation requires a costly…
- arXiv cs.AIagentsThe Harness Effect: How Orchestration Design Sets the Token Economics of Enterprise Agentic AI
arXiv:2607.06906v2 Announce Type: replace Abstract: Agentic AI development today runs on token maxing: buying capability with tokens -- longer reasoning traces, more turns, wider tool payloads, bigger replayed contexts -…
- arXiv cs.AIresearchMeasuring Intelligence Beyond Human Scale
arXiv:2607.07040v2 Announce Type: replace Abstract: How can we measure intelligence beyond human capability? Human-authored benchmarks saturate, and above human capability, examiners may not know which tasks are both har…
- arXiv cs.AIresearchWhat Keeps Vision-Language Models Looking at the Image?
arXiv:2607.12815v3 Announce Type: replace Abstract: When do vision-language models need direct access to the image while generating an answer? We study image dependence during answer generation by examining how the visua…
- arXiv cs.AIagentsCura 1T: Healthcare Foundation Model via Recursive Self-Improvement
arXiv:2607.15314v3 Announce Type: replace Abstract: Healthcare spans high-stakes communication, expert reasoning, and workflow execution, yet specialized language models that cover these use cases together remain limited…
- arXiv cs.AIresearchA Dual-Hypothesis Reasoning Framework for LLM Guardrails
arXiv:2607.17575v3 Announce Type: replace Abstract: We propose ARBITER, a novel LLM guardrail framework that introduces two key ideas: (i) dual-hypothesis reasoning, a reasoning method for LLM guardrails that explicitly…
- arXiv cs.AIresearchNeuro-Symbolic Meta-Policies for Temporal Knowledge-Graph Memory under Partial Observability
arXiv:2607.18368v5 Announce Type: replace Abstract: Partially observable reinforcement learning requires deciding what to retain, retrieve, and forget over time. We introduce a neuro-symbolic meta-policy that learns whic…
- arXiv cs.AIresearchTILT: Model-Intrinsic Reward Alignment For Compositional Diffusion
arXiv:2607.21606v2 Announce Type: replace Abstract: Consider conditional generation $p(x \mid C=\{c_1, c_2, \dots c_k\})$ where $C$ is a prompt composed of multiple concepts $c_i$. Diffusion models often struggle with co…
- arXiv cs.AIresearchDeconstructing Off-Policy Ratios: Entropy-Normalized Trust Regions for Asynchronous Reinforcement Learning
arXiv:2607.22186v5 Announce Type: replace Abstract: Asynchronous reinforcement learning (RL) accelerates large language model (LLM) post-training by overlapping rollout generation with policy optimization, but the result…
- arXiv cs.AIresearchTowards Trustworthy Physical Intelligence: From Theory to Practice Across Life Cycle
arXiv:2607.22877v4 Announce Type: replace Abstract: Physical intelligence refers to intelligence systems that understand, reason about, and act in accordance with the physical world and its underlying laws, dynamics, and…
- arXiv cs.AIresearchDimensionality and Measurement Precision in HLE's Multiple-Choice Subset
arXiv:2607.27420v2 Announce Type: replace Abstract: Humanity's Last Exam (HLE) is widely used to evaluate frontier language models. HLE organizes its questions into eight subject-domain categories, whose subscores are of…
- arXiv cs.AIagentsSKILL-KD: Contrastive Skill Distillation for LLM Agents
arXiv:2607.28048v4 Announce Type: replace Abstract: Skill-based prompting has become a practical mechanism for improving LLM agents, yet existing methods often treat skills as summaries of the agent's own experience or o…
- arXiv cs.AIresearchRA-CAD: Learning Post-Execution Critique for State-Aware Text-to-CAD Generation
arXiv:2608.05714v4 Announce Type: replace Abstract: Text-to-CAD generation translates natural-language design intent into editable and executable parametric computer-aided design (CAD) codes, reducing the expertise and e…
- arXiv cs.AIresearchNxN E-valuation: Hypothesis Certification via a Conformal CRT Null
arXiv:2608.06621v5 Announce Type: replace Abstract: Language models now propose hypotheses faster than anyone can check them, so the step that certifies them must need nothing from the proposer but the hypothesis. We sta…
- arXiv cs.AIresearchThe Value of a Prompt: An LLM-Relative Kolmogorov-Complexity Approach
arXiv:2608.16438v2 Announce Type: replace Abstract: In a world where valuable artifacts are increasingly created, completed, or processed by LLMs, the central economic question is not only what the LLM can produce, but w…
- arXiv cs.AIagentsCTIFoundry: An Agent-Native Corpus Scaffold for Cyber Threat Intelligence
arXiv:2608.18613v2 Announce Type: replace Abstract: Cyber threat intelligence (CTI) is increasingly consumed not by human analysts but by LLM agents that compose multi-step investigations at query time. The harness side…
- arXiv cs.AIresearchHidden in the Request: Explaining Unethical LLM Compliance through Token Relevance
arXiv:2608.23264v3 Announce Type: replace Abstract: Although Large Language Models (LLMs) are aligned to optimize for both helpfulness and harmlessness, these dual objectives may conflict, inevitably leading to alignment…
- arXiv cs.AIresearchTowards Reliable, Generalizable, and Specific In-Context Knowledge Editing via Multi-Objective Reinforcement Learning
arXiv:2608.25100v3 Announce Type: replace Abstract: Large Language Models (LLMs) are powerful but limited by static parametric knowledge that becomes outdated once pretraining ends. Knowledge editing addresses this probl…
- arXiv cs.AIagentsRealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requests
arXiv:2608.27831v3 Announce Type: replace Abstract: Coding agents are now commonly evaluated on the SWE-bench family of benchmarks, whose tasks are built from curated GitHub issues: long, structured, and information-rich…
- arXiv cs.AIresearchSEPO: Evidence-Grounded Prompt Optimization via Structural Editing
arXiv:2608.28067v2 Announce Type: replace Abstract: Existing API-only prompt optimisers are often described as interpretable, but in practice, this usually means only post-hoc inspectability: each iteration still rewrite…
- arXiv cs.AIresearchDramaChain Bench: An End-to-End Benchmark for Short-Drama Generation
arXiv:2609.00646v2 Announce Type: replace Abstract: Commercial short-drama production follows a multi-stage chain: script, storyboard, keyframe imagery, shot-level video, and the finished short drama. Most existing bench…
- arXiv cs.AIresearchCan LLMs Discover Scientific Laws in Real and Parallel Worlds?
arXiv:2609.01552v2 Announce Type: replace Abstract: Scientific law discovery has long been central to scientific progress, proceeding through iterative cycles of generating hypotheses, testing them against empirical evid…
- arXiv cs.AIresearchAnswer-Distribution Trajectories: A Stochastic-Dynamics View of LLM Reasoning
arXiv:2609.09030v2 Announce Type: replace Abstract: Chain-of-thought reasoning provides a structured computation between a model's input and final answer. Yet it is often evaluated through endpoint accuracy, which ignore…
- arXiv cs.AIresearchA Function-Space Approach to the Statistical Mechanics of Learning Dynamics
arXiv:2609.09589v2 Announce Type: replace Abstract: In the kernel regime, neural-network learning inherits its preferences from a frozen spectrum. During feature learning, this spectrum evolves, yet networks retain syste…
- arXiv cs.AIresearchReading the Whole Heart: Latent-Attention Masked Autoencoders for Multimodal Cardiac Representation Learning
arXiv:2609.12035v3 Announce Type: replace Abstract: Cardiovascular diagnosis and treatment rest on integrating complementary modalities, such as electrocardiogram, echocardiography, and chest X-rays, each capturing disti…
- arXiv cs.AIresearchSchizophrenia Detection from EEG Signals: A Transformer Framework with Spectrogram Representation
arXiv:2609.14015v2 Announce Type: replace Abstract: Schizophrenia is a serious psychiatric disorder that affects millions of people worldwide, and its diagnosis remains primarily dependent on clinical assessment. Electro…
- arXiv cs.AIresearchUnderstanding Certified Training with Interval Bound Propagation
arXiv:2306.10426v3 Announce Type: replace-cross Abstract: As robustness verification methods are becoming more precise, training certifiably robust neural networks is becoming ever more relevant. To this end, certified t…
- arXiv cs.AIresearchPPFedIT: Towards Privacy-Preserving Federated Instruction Tuning with Few-shot Local Examples
arXiv:2403.06131v3 Announce Type: replace-cross Abstract: Instruction tuning aligns large language models (LLMs) with human intentions but requires diverse, high-quality data that are difficult to collect in privacy-sens…
- arXiv cs.AIresearchCHAOSMINING: Benchmarking Post-Hoc Attribution with Sparse Informative Features in High Dimensions
arXiv:2406.12150v2 Announce Type: replace-cross Abstract: Post-hoc attribution is widely used to identify important model inputs, but evaluating whether these attributions identify truly informative features is difficult…