AI news archive.
19,861 articles, newest first, from the outlets listed on the sources page.
- arXiv cs.AIagentsEvidence-Inference Reconstruction: When The Evidence Is Recalled But The Reasoning Goes Wrong
arXiv:2609.33778v1 Announce Type: new Abstract: Modern multi-hop LLM agents are equipped with built-in mechanisms to detect errors in intermediate reasoning steps. Such errors trigger corrective actions from these agents…
- arXiv cs.AIresearchIs your uncertainty map wrong, or is its target? Exact diagnostics for the Tweedie diagonal, and a gradient-free alternative
arXiv:2609.33786v1 Announce Type: new Abstract: A diffusion model can predict a follow-up medical scan from a baseline, but a clinician needs a per-voxel map of where that prediction can be trusted. Many such maps approx…
- arXiv cs.AIresearchDual-Vocabulary Language Model for Cross-Tokenizer Distillation
arXiv:2609.33816v1 Announce Type: new Abstract: On-policy distillation (OPD) bridges teacher supervision and student behavior, but different teacher-student tokenizers introduce misalignment in both input tokenization (#…
- arXiv cs.AIagentsVestrum: Improving Agent Harnesses by Adapting Their Verification, Structure and Memory
arXiv:2609.33822v1 Announce Type: new Abstract: An agent harness controls how a language model accesses information, uses tools, preserves memory, and checks its work. Improving this software is costly when each evaluati…
- arXiv cs.AIagentsLaya as a Typed Probabilistic Assessor: An Independent Reproduction and a Preregistered Study of Calibration and Selective Escalation
arXiv:2609.33843v1 Announce Type: new Abstract: The shipped Laya Typed-Decisions checkpoint, a 421M-parameter ModernBERT-large assessor that answers typed choice/noul/score questions over workflow state, is uniformly und…
- arXiv cs.AIresearchHow code helps different tasks? A decompositional lens on LLM post-training
arXiv:2609.33845v1 Announce Type: new Abstract: Evaluating code data as a single corpus can obscure which types of code data benefit which models and downstream tasks. Effective data selection requires understanding both…
- arXiv cs.AIagentsR$^2$ Flow: Recursive Self-Improvement via Recursive Skill Evolution
arXiv:2609.33867v3 Announce Type: replace Abstract: LLM-based agents can improve themselves across tasks by reusing and revising the skills they orchestrate into executable procedures. Flow-based training fits this loop:…
- arXiv cs.AIagentsWhen Successful Strategies Fail: Adaptation to Environmental Novelty in Terminal Agents
arXiv:2609.33870v1 Announce Type: new Abstract: LLM agents increasingly solve long-horizon tasks by autonomously interacting with their environment. In doing so, their strategies rely on assumptions about that environmen…
- arXiv cs.AIresearchCurating Merchant-Matching Training Data with Two Confidence-Gated Local LLM Judges
arXiv:2609.33878v2 Announce Type: replace Abstract: Merchant matching resolves a noisy payment descriptor to a retrieved merchant entity or returns no match. A key challenge in curating training labels is distinguishing…
- arXiv cs.AIagentsWhen Consent Outlives Context: Residual Authority Replay in Long-Lived Agents
arXiv:2609.33910v1 Announce Type: new Abstract: LLM agents increasingly rely on user approval to authorize security-sensitive actions at runtime. Such approvals are granted within a specific task and execution context. I…
- arXiv cs.AIagentsHyperMCTS: Hypergraph-Augmented MCTS for Long-Horizon LLM Agents
arXiv:2609.33920v1 Announce Type: new Abstract: Long-horizon tasks require large language model (LLM) agents to coordinate decisions under constraints that span an entire solution. Monte Carlo Tree Search (MCTS) offers a…
- arXiv cs.AIagentsDesigning Reliable LLM-as-a-Judge Measurement Systems for Multi-Turn Business Agents
arXiv:2609.33955v1 Announce Type: new Abstract: Many LLM-as-a-judge evaluations score fixed outputs under a fixed task definition. Production multi-turn business agents instead require a maintained measurement system: co…
- arXiv cs.AIresearchEHRAdapt: Adapting Pretrained Language Models to Electronic Health Records with Semantic Priors for Rare Clinical Events
arXiv:2609.34007v2 Announce Type: replace Abstract: Electronic health records (EHRs) encode clinical histories as (time, modality, code) tuples, whereas pretrained language models expect text tokens. Serializing tuples a…
- arXiv cs.AIresearchA Computer Vision Approach to Visual Fraud Detection in Phishing Websites Using YOLOv8
arXiv:2609.34015v2 Announce Type: replace Abstract: Phishing remains one of the most common vectors for financial and identity fraud, and most detection systems still rely on inspecting a page's URL, HTML markup, or doma…
- arXiv cs.AIresearchJev in Medicine: A Benchmark Evaluation
arXiv:2609.34024v4 Announce Type: replace Abstract: Jev is a non-generative System One model that assigns probabilities to predefined answer options and cannot answer outside them. Its accuracy and calibration on medical…
- arXiv cs.AIagentsLarge Language Models for Structured Clinical Data Analysis: Dual-Agent Grounding and Validation
arXiv:2609.34039v1 Announce Type: new Abstract: Objective: To develop and characterize CLEAR-Med, a dual-agent framework for natural-language analysis of structured clinical data that separates SQL-based invocation from…
- arXiv cs.AIresearchThinking Outside the Box: Retention and Transmission of Information in Sliding-Window KV Inference
arXiv:2609.34049v1 Announce Type: new Abstract: Sliding-window KV inference refers to processing a sequence incrementally while retaining only a fixed-size cache of recent key and value states. It can be applied to pretr…
- arXiv cs.AIagentsTowards Certificate-Driven Software Porting: A Self-Improving Agentic Harness for Scientific Program Optimization
arXiv:2609.34069v1 Announce Type: new Abstract: The upgrade and rewriting of large scientific codebases has traditionally been a major challenge. While evolutionary search with large language models (LLMs) can port and a…
- arXiv cs.AIagentsPhysFieldBench: Can Multimodal Models Understand Physical Fields?
arXiv:2609.34072v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) are increasingly envisioned as core components of scientific and engineering agents, yet their ability to interpret physical fields…
- arXiv cs.AIresearchGenoMorph: Pathway-Grounded Genomic Disease Reasoning via Adaptive Latent Computation
arXiv:2609.34079v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated strong capabilities in biological reasoning; however, genomic disease inference remains largely dependent on memorized gene-d…
- arXiv cs.AIresearchK-OPSD: Verifiable On-Policy Self-Distillation for Post-Training Vision-Language Models on AEC Drawings
arXiv:2609.34082v1 Announce Type: new Abstract: Interpreting architecture, engineering, and construction (AEC) drawings is hard for general Multimodal Large Language Models (MLLMs) and vision-language models (VLMs). We i…
- arXiv cs.AIresearchA Differentiable Optimization Framework for Registering Sequential Bounding Boxes with Point Cloud Stream
arXiv:2609.34103v1 Announce Type: new Abstract: Refining a sequence of coarse 3D bounding boxes against a LiDAR point-cloud stream demands tracks that are geometrically accurate (high IoU) and temporally coherent (low ro…
- arXiv cs.AIresearchSpecRegMatch: Robust Semi-Supervised Regression for Vehicle Interior Noise Prediction
arXiv:2609.34111v1 Announce Type: new Abstract: The rapid advancement of artificial intelligence has observed increased application in predicting vehicle interior noise levels within the automotive industry. However, the…
- arXiv cs.AIagentsGUITAR: Structured Failure Diagnosis of GUI Agents via State Transitions
arXiv:2609.34113v2 Announce Type: replace Abstract: Understanding where and why Graphical User Interface (GUI) agents fail is essential for building more reliable systems, yet current evaluation relies on step accuracy,…
- arXiv cs.AIagentsFrom Attack Success to Attack Severity: Counterfactual Memory Attacks on LLM Agents
arXiv:2609.34132v1 Announce Type: new Abstract: As LLM agents increasingly rely on persistent memory for long-horizon and personalized behavior, they can retain and reuse information across interactions, but this also cr…
- arXiv cs.AIagentsStateGuard: Analytical-State Management with Validity-Aware Intervention for Long-Horizon Data Agents
arXiv:2609.34134v1 Announce Type: new Abstract: LLM-based agents have shown strong capabilities in automated data analysis and are increasingly moving toward long-horizon, multi-stage analytical workflows. However, as th…
- arXiv cs.AIagentsEvo2Team: When Do Evolved Skills Transfer? From Selection to Deployment
arXiv:2609.34135v1 Announce Type: new Abstract: A skill bank that helps one multi-agent system may leave another's behavior unchanged. A transferred rule helps only when target agents act on it successfully. We study thi…
- arXiv cs.AIagentsWaggle: Learning One Anonymous Local Law for Self-Organizing LLM Swarms
arXiv:2609.34136v1 Announce Type: new Abstract: As LLM agents increasingly collaborate on complex tasks, how to organize their interactions becomes a central design question. Existing multi-agent systems typically learn…
- arXiv cs.AIresearchYou Can't Have It Both Ways: Concept Entanglement Limits Diffusion Model Unlearning
arXiv:2609.34137v1 Announce Type: new Abstract: Concept unlearning in text-to-image diffusion models aims to suppress a target concept (e.g., \texttt{horse}) while preserving related but distinct content (e.g., \texttt{d…
- arXiv cs.AIagentsSame Tasks, Different Apps: Why Mobile GUI Agents Fail to Generalize?
arXiv:2609.34139v1 Announce Type: new Abstract: Mobile GUI agents deployed in real settings must work across different applications that support the same functionality. Most existing benchmarks test each task in only one…
- arXiv cs.AIagentsSelf-Evolving Agents via Likelihood-Guided Tool-Space Optimization
arXiv:2609.34151v1 Announce Type: new Abstract: Self-evolving agents can continually improve their behavior, while tools define the executable action space through which they interact with the environment. However, expos…
- arXiv cs.AIagentsTableSeek: Structure-Preserving Agentic Evidence Seeking over Heterogeneous Table Corpora
arXiv:2609.34157v1 Announce Type: new Abstract: Open-domain table retrieval seeks tables that contain sufficient evidence for answering a question or verifying a claim. Yet semantic relevance is often misleading: topical…
- arXiv cs.AIagentsRoutePrism: Tracing Construction Order Effects in Agent Memory
arXiv:2609.34160v1 Announce Type: new Abstract: Processing the same records in a different order can discard different evidence, yet endpoint accuracy alone cannot reveal what changed or whether it mattered. We introduce…
- arXiv cs.AIagentsReplayLens: Auditing Agents' Use of Outcomes
arXiv:2609.34177v2 Announce Type: replace Abstract: When an agent reuses logged experience, a changed decision may reflect the recorded score, the action's name, or the record's position in storage. Standard memory evalu…
- arXiv cs.AIresearchRAGWarrant: Evidence-Preserving Governance for RAG Policy Promotion Under Quality, Cost, Latency, and Risk Constraints
arXiv:2609.34179v1 Announce Type: new Abstract: Retrieval-augmented generation systems are extensively instrumented with metrics, benchmarks, traces, and automated judges, but these tools do not decide whether a proposed…
- arXiv cs.AIresearchDecision Readouts for Text-Mediated Video Anomaly Detection: An Exploratory Evaluation of Jev and Qwen
arXiv:2609.34180v1 Announce Type: new Abstract: How much does the decision readout matter when video-derived textual evidence is held fixed? We evaluate Jev typed decisions and three Qwen readouts on a sparse development…
- arXiv cs.AIresearchEfficient Reasoning via Constrained Optimization in Latent Space
arXiv:2609.34181v2 Announce Type: replace Abstract: Large Reasoning Models (LRMs) have shown remarkable reasoning capabilities, yet they still suffer from overthinking, generating redundant reasoning steps which incur su…
- arXiv cs.AIresearchCASS: Contribution-Aware Structured Sparsity for Model Merging
arXiv:2609.34184v1 Announce Type: new Abstract: Model merging integrates task-specific fine-tuned models into a single multi-task model, but often suffers from parameter interference caused by conflicting task-vector upd…
- arXiv cs.AIresearchPainterBench: A Figural Divergent-Thinking Benchmark for Tool-Using Language Models
arXiv:2609.34195v1 Announce Type: new Abstract: Figural divergent thinking is the ability to develop a given shape fragment into an original drawing. In humans, this ability is assessed with incomplete-drawing tasks. We…
- arXiv cs.AIresearchBehavior-Grounded Semantic Enrichment for Financial Fraud Modeling and Reasoning
arXiv:2609.34211v2 Announce Type: replace Abstract: In financial fraud detection, rich semantic context can provide important evidence for transaction behavior modeling and fraud reasoning. However, public real-world fin…
- arXiv cs.AIagentsGlyphBench: A Playground for Language-Model Reinforcement Learning
arXiv:2609.34214v1 Announce Type: new Abstract: We introduce GlyphBench, an environment suite for reinforcement learning (RL) post-training of language-model agents, with over 360 tasks spanning diverse games. GlyphBench…
- arXiv cs.AIagentsSame Winners, Different Success Rates: Evaluating How LLM Agents Recover from Failures
arXiv:2609.34215v1 Announce Type: new Abstract: Evaluating how LLM agents recover from mid-task failures is central to deploying reliable agentic systems. Existing checkpoint-based benchmarks measure recovery by comparin…
- arXiv cs.AIagentsWhen Does Selection Replace Extraction? A Pre-Registered Test of Agent Memory with a Typed Decision Model
arXiv:2609.34227v1 Announce Type: new Abstract: Does conversational memory need LLM-extracted facts, or is selecting the right raw turns enough? Published results disagree. Extraction-based systems report gains from dist…
- arXiv cs.AIagentsAdaGuard: An Adaptive Guard Model with User-defined Policies
arXiv:2609.34241v2 Announce Type: replace Abstract: Guard models support the safe deployment of language model agents, but fixed risk taxonomies limit their ability to accommodate requirements that vary across applicatio…
- arXiv cs.AIagentsStashbird: Episode-Grounded Lifecycle Reconciliation for Conversational Agent Memory
arXiv:2609.34242v2 Announce Type: replace Abstract: AI agents require memory that remains consistent as conversational evidence changes. We present Stashbird, an episode-grounded memory system that uses provenance to rec…
- arXiv cs.AIresearchEvolving Support Priorities in Empathetic Reinforcement Learning
arXiv:2609.34249v2 Announce Type: replace Abstract: We identify a fundamental mismatch in empathetic reinforcement learning: support priorities evolve with the dialogue state, yet existing methods typically optimize pred…
- arXiv cs.AIresearchQuantaSpike: Short-Window Spike-Driven Quantization for Large Language Models
arXiv:2609.34259v1 Announce Type: new Abstract: Large language models (LLMs) achieve strong performance across many tasks but rely on dense multiply-accumulate (MAC) operations during inference, resulting in high energy…
- arXiv cs.AIagentsMaintaining Benchmarks Against Increasingly Capable Agents: Detection and Remediation of Unearned Passes
arXiv:2609.34262v1 Announce Type: new Abstract: Agentic benchmarks guide model selection and training. Yet an agent can pass a task without demonstrating the intended capability. Such outcomes constitute unearned passes;…
- arXiv cs.AIresearchQuery Expansion and Key Specialization in Transformer Attention Geometry
arXiv:2609.34273v1 Announce Type: new Abstract: The projection of queries and keys are central to the attention mechanism in Transformer architectures. While they are mathematically symmetric, they play different roles i…
- arXiv cs.AIagentsBIABench: Evaluating AI agents on real-world bioimage analysis tasks
arXiv:2609.34274v1 Announce Type: new Abstract: Artificial-intelligence (AI) agents hold promise for automating bioimage analysis, yet no benchmark evaluates whether they can carry out real-world analyses end to end. Suc…