AI news archive.
10,659 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchWAKMan: Workload Aware KVCache Manager
arXiv:2610.36355v1 Announce Type: new Abstract: Prefix caching enables a Large Language Model (LLM) servers to reuse the Key-Value (KV) states and accelerate prefill computation stage for requests that share a prefix wit…
- arXiv cs.AIresearchLoRA on a Budget: Function-Aware Post-Hoc Compression for Adapter Fleets
arXiv:2610.36386v1 Announce Type: new Abstract: Low-Rank Adaptation (LoRA) has become a key mechanism for efficiently specializing foundation models into many task- and user-specific variants. Serving many LoRA adapters…
- arXiv cs.AIresearchkw-bench: Breaking Frontier Harnesses on Real-World Tasks
arXiv:2610.36405v1 Announce Type: new Abstract: Each new generation of frontier model pushes the edge of what is possible. Inside a harness -- a loop, a shell, a filesystem, the ability to write and run its own code -- a…
- arXiv cs.AIresearchVerifiable Efficiency Limits of Innate LLM Planning in Sliding Puzzles
arXiv:2610.36415v1 Announce Type: new Abstract: Tool-assisted planning success can reflect external search and verification, leaving a model's unaided planning ability unresolved. We isolate innate planning through compl…
- arXiv cs.AIresearchFrom Solver Trajectories to Teaching Trajectories: Cognition-Aligned Reasoning Distillation
arXiv:2610.36445v1 Announce Type: new Abstract: Knowledge distillation from large reasoning models (LRMs) often uses raw teacher trajectories as supervision. However, this practice creates a mismatch between the goal of…
- arXiv cs.AIresearchA Heart Means Recursion: Exploring Visual Meaning and Communication in VLMs
arXiv:2610.36449v1 Announce Type: new Abstract: Do vision-language models (VLMs) share our ability to communicate visually? We share preliminary findings, starting from a surprising observation: simple visual symbols lik…
- arXiv cs.AIresearchAC-JEPA: Anchored Cross-Modal Joint Embedding Predictive Architecture for Visual-to-Symbolic Translation
arXiv:2610.36493v1 Announce Type: new Abstract: Cross-modal translation requires models to capture shared semantics across modalities while preserving the fine-grained structure required by the target modality. Yet in re…
- arXiv cs.AIresearchMedConstBench: Benchmarking How Constitutions Steer LLMs in Medical Ethics
arXiv:2610.36513v1 Announce Type: new Abstract: As large language models advance rapidly, they are increasingly used to answer medical questions, and many such models are now deployed locally. Once a model is deployed, i…
- arXiv cs.AIresearchATEMPO: Amortized Transfer Entropy Model Per-lag Over-Time
arXiv:2610.36536v1 Announce Type: new Abstract: When one time series informs another, the delay at which that information arrives determines if models can utilize this relationship effectively. A source may be strongly p…
- arXiv cs.AIresearchADTGuard: Adaptive Depth-Time Safeguard for Streaming Inference in Large Language Models
arXiv:2610.36574v1 Announce Type: new Abstract: Large Language Models (LLMs) can generate fluent responses to diverse user requests, yet streaming may expose users to harmful content before generation is complete. Existi…
- arXiv cs.AIresearchFollow the Herd: A Data-Driven Approach for Actionable Recourse
arXiv:2610.36580v1 Announce Type: new Abstract: As machine learning systems make increasingly greater decisions about people's lives, it is critical to provide users with actionable recourse about how they can achieve th…
- arXiv cs.AIresearchLearning to Formalize: Bridging Open-Ended Workloads and Fast Decision Models
arXiv:2610.36595v1 Announce Type: new Abstract: Decision models offer fast choices over sup- plied alternatives, while general-purpose large language models interpret open-ended tasks through flexible generation and reas…
- arXiv cs.AIresearchPolarity--Time Interaction and Multi-Level Reliability for Event-Based Spacecraft Pose Estimation
arXiv:2610.36625v1 Announce Type: new Abstract: Sparse event responses produce uneven keypoint errors whose effects on spacecraft pose depend on correspondence geometry and the reliability of successive observations. We…
- arXiv cs.AIresearchThe uneven rise of AI-generated web content
arXiv:2610.36631v1 Announce Type: new Abstract: In less than five years, generative artificial intelligence (AI) has dramatically reshaped the public-facing web. Using a sample of more than 5 million crawlable webpages,…
- arXiv cs.AIresearchLearning Where to Trust: Reliability-Guided Expert Routing and Velocity Gating for 3D Deformable Registration
arXiv:2610.36672v1 Announce Type: new Abstract: Deformable image registration relies on accurate anatomical correspondences. However, explicitly coupling local matching reliability with spatial context selection and defo…
- arXiv cs.AIresearchPsyJev: Psychological Analysis via System-1 Judgment
arXiv:2610.36689v1 Announce Type: new Abstract: Self-disclosures on the Web support diverse psychological analyses, from recognizing expressed experiences to identifying symptom-relevant information. We study a shared, q…
- arXiv cs.AIresearchCan LLMs Reliably Grade Olympiad Proofs? A Study of Mathematical Verification with LLMs
arXiv:2610.36767v1 Announce Type: new Abstract: Large Language Models (LLMs) have made remarkable progress on solving competition-level mathematics, yet their ability to verify natural-language mathematical proofs remain…
- arXiv cs.AIresearchDependency Graphs of Mathematical Arguments: Representation, Structural Measurement, and Proof Review
arXiv:2610.36776v1 Announce Type: new Abstract: We propose a source-grounded framework for representing and measuring mathematical arguments with assistance from large language models. It includes numbered and unnumbered…
- arXiv cs.AIresearchAn Incremental Knowledge-Graph Learner Digital Twin for Evidence-Grounded Early Warning
arXiv:2610.36820v2 Announce Type: replace Abstract: Early-warning models in learning analytics can identify learners at risk of failure or withdrawal, but the resulting alerts are often only weakly connected to inspectab…
- arXiv cs.AIresearchConservative Policy Iteration For Zero-Sum POSGs
arXiv:2610.36847v1 Announce Type: new Abstract: Occupancy-state reductions make dynamic programming possible for zero-sum partially observable stochastic games (zs-POSGs). Existing point-based solvers usually optimise a…
- arXiv cs.AIresearchNoise Steering: Exploiting Wake Steering and Acoustic Directivity for Noise-Constrained Wind Farm Control via Reinforcement Learning
arXiv:2610.36888v2 Announce Type: replace Abstract: Existing noise mitigation strategies for wind farms usually consist in reducing turbines power output, without accounting for the wake interactions between them and the…
- arXiv cs.AIresearchNatural Hard-Negative Mining from Positives for Instruction-Following Retrieval
arXiv:2610.36900v1 Announce Type: new Abstract: Instruction-following information retrieval (IFIR) requires retrievers to rank documents by both a query and its instance-specific instructions, yet instruction-conditioned…
- arXiv cs.AIresearchDo bots feel bad? Assessing empirical evidence for functional valence in LLMs
arXiv:2610.36962v1 Announce Type: new Abstract: Do LLMs have valenced states, i.e. states that feel pleasant or unpleasant? Recent empirical work has claimed to show that LLMs possess functional valenced states (e.g. ``f…
- arXiv cs.AIresearchSpectRAG: Weighted Chunk Graphs for Token-Budgeted Context Selection in Multi-Hop Retrieval-Augmented Generation
arXiv:2610.36991v3 Announce Type: replace Abstract: Retrieval-augmented generation (RAG) systems ground large language models (LLMs) by retrieving external evidence and conditioning generation on the retrieved context. A…
- arXiv cs.AIresearchGuided-RLVR: Guided Exploration in Reinforcement Learning with Verifiable Rewards
arXiv:2610.37041v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has shown strong empirical gains in improving the reasoning ability of large language models (LLMs) on tasks such as m…
- arXiv cs.AIresearchAI-Assisted Discovery of CNOT Decompositions for Quantum Permutations on Constrained Topologies
arXiv:2610.37045v1 Announce Type: new Abstract: Efficient quantum circuit compilation is challenging for quantum computers, where limited qubit connectivity can introduce substantial overhead in the implementation of log…
- arXiv cs.AIresearchCorrectness Conditioning as Selection in Chain-of-Thought Faithfulness Evaluation
arXiv:2610.37070v1 Announce Type: new Abstract: Restricting chain-of-thought (CoT) faithfulness evaluation to correctly answered instances changes the population being measured. We study this selection step in a seven-va…
- arXiv cs.AIresearchProspector: learning a compact graph traversal policy for scientific reasoning
arXiv:2610.37121v1 Announce Type: new Abstract: Scientific reasoning in specialized engineering fields depends on knowledge scattered across the literature, often proprietary and absent from frontier LLM training data. I…
- arXiv cs.AIresearchSFE-O: An Ontology for Knowledge-Grounded AI in Supercritical CO$_2$ Biomass Extraction and Fractionation
arXiv:2610.37123v1 Announce Type: new Abstract: Supercritical CO$_2$ extraction (CO$_2$-SFE) offers a sustainable route for recovering high-value compounds from biomass while limiting the use of conventional organic solv…
- arXiv cs.AIresearchR\`enyi-Regularized Flow-Based Variational Inference for Bayesian Calibration
arXiv:2610.37169v1 Announce Type: new Abstract: Variational inference with normalizing flows provides a flexible posterior approximation for Bayesian model calibration, but struggles when the posterior has several modes…
- arXiv cs.AIresearchGraphRAGProv: GraphRAG Meets How-Provenance
arXiv:2610.37237v1 Announce Type: new Abstract: GraphRAG systems often use knowledge graphs to retrieve relevant context for an LLM that produces the final answer. That makes answers more inscrutable and therefore less a…
- arXiv cs.AIresearchTCM-DiagCommit: A TCM-SD-Derived Benchmark for Evidence-Conditioned Syndrome Selection and Commitment
arXiv:2610.37243v1 Announce Type: new Abstract: Traditional Chinese Medicine (TCM) syndrome differentiation is usually evaluated as static final-label selection over complete clinical records. This setting hides three di…
- arXiv cs.AIresearchRPSD: Retrieval-Privileged On-Policy Self-Distillation with Optimal Transport
arXiv:2610.37281v1 Announce Type: new Abstract: Sequential recommendation predicts a user's next interaction from historical behavior. Recent LLM-based methods recast this task as generative recommendation and increasing…
- arXiv cs.AIresearchReForest: Interpretability Meets Reasoning in Isolation Forest
arXiv:2610.37316v1 Announce Type: new Abstract: Statistical anomaly detection methods like Isolation Forest are efficient but suffer from a critical semantic gap. They identify numerical rarities but fail to understand t…
- arXiv cs.AIresearchUncovering Why LLMs Fail: From Internal Mechanisms to Adversarial Reasoning Trajectories
arXiv:2610.37369v2 Announce Type: replace Abstract: Large language models (LLMs) are usually evaluated by final-output correctness or safety, yet an endpoint label does not reveal how a failure develops throughout the re…
- arXiv cs.AIresearchBeyond the Readout: A Latent Numerical Frame Shared Across Modality and Operation
arXiv:2610.37422v1 Announce Type: new Abstract: When a multimodal model computes 17+35 from text or 91-39 from a rendered image, it produces the same answer, 52. Does it use one internal concept of number for both, or se…
- arXiv cs.AIresearchTimestral
arXiv:2610.37585v1 Announce Type: new Abstract: We introduce Timestral, a family of four time series foundation models for probabilistic forecasting, with 5M, 28M, 123M, and 209M parameters. Timestral sets a new state of…
- arXiv cs.AIresearchRare, Relevant, and Still Wrong: How Memory Fails LLM Group Decisions
arXiv:2610.37624v1 Announce Type: new Abstract: When an LLM group's discussion is compressed into a short memory for a final decision maker, what should the memory keep? Two natural heuristics keep the rarest claims or t…
- arXiv cs.AIresearchWhat Improves During Pretraining? Tracking Knowledge Acquisition with Point-in-Time Models
arXiv:2610.37631v1 Announce Type: new Abstract: Language models pretrained on web corpora collected over multiple years learn both general and domain-specific knowledge. How these different kinds of knowledge are learned…
- arXiv cs.AIresearchKPPax: preprocess differentiable kinetic ODE for machine learning in atmospheric chemistry
arXiv:2610.37655v1 Announce Type: new Abstract: Atmospheric chemical mechanisms are commonly distributed in the language of the Kinetic PreProcessor (KPP), which generates efficient Fortran or C code but does not fit int…
- arXiv cs.AIresearchASTROLABE: A 3D Particle-Level Surrogate for Cosmology Simulation with Physical Flow Matching
arXiv:2610.27068v1 Announce Type: cross Abstract: The microphysics of dark matter (DM) is imprinted on cosmological structure across an enormous dynamic range. Nonlinear cosmological observables are most reliably modeled…
- arXiv cs.AIresearchscPyviewer: a Python-native interactive viewer from AnnData single-cell data
arXiv:2610.35890v1 Announce Type: cross Abstract: Motivation: Interactive web tools enabling non-programmers to explore single-cell datasets, embeddings, gene expression, metadata, and marker genes are standard laborator…
- arXiv cs.AIresearchBeyond Greenwashing: Algorithmic Resilience and the Non-Linear Governance Premium in Sovereign Capital Flows
arXiv:2610.35895v1 Announce Type: cross Abstract: Despite the proliferation of sovereign Environmental, Social, and Governance (ESG) metrics, traditional linear macro-financial models systematically misprice Foreign Dire…
- arXiv cs.AIresearchTask-Utility IQA: Rethinking Image Quality Assessment in Embodied Tasks
arXiv:2610.35905v1 Announce Type: cross Abstract: Image quality assessment (IQA) underpins embodied imaging pipelines by judging whether visual quality satisfies downstream tasks, yet most existing methods learn task-agn…
- arXiv cs.AIresearchPhysiognomy Revisited: How Vision-Language Models Encode Sensitive Attributes Through Image Transformation
arXiv:2610.35910v1 Announce Type: cross Abstract: Emergent generative AI technologies, while remarkably beneficial across several tasks and domains, can suffer from unforeseen vulnerabilities and difficult-to-anticipate…
- arXiv cs.AIresearchReCoPII: A Relation-Grounded Framework for Building Contextual PII Masking Datasets
arXiv:2610.35925v1 Announce Type: cross Abstract: Training contextual masking systems for personally identifiable information (PII) requires examples distinguishing maskable from preservable mentions of one entity type,…
- arXiv cs.AIresearchResCom: Activation Residual Compensation for Large Language Model Quantization
arXiv:2610.35955v1 Announce Type: cross Abstract: Low-bit weight--activation quantization reduces the memory and computation needed to deploy large language models, but activation rounding can still degrade accuracy afte…
- arXiv cs.AIresearchStructured Time-Frequency Representation for Channel Prediction
arXiv:2610.35956v1 Announce Type: cross Abstract: Accurate CSI forecasting is essential for mitigating channel aging and supporting reliable downlink transmission in high-mobility wireless systems. In FDD systems, foreca…
- arXiv cs.AIresearchMultiple Leak Localization in Water Distribution Networks through Graph Neural Networks
arXiv:2610.35972v1 Announce Type: cross Abstract: Leak localization in water distribution networks is challenging when multiple leaks occur simultaneously, particularly under sparse monitoring. Existing data-driven metho…
- arXiv cs.AIresearchLet the Gate Decide: Per-Lookup Precision for Quantized Conditional Memory in LLMs
arXiv:2610.35985v1 Announce Type: cross Abstract: Conditional memory layers, such as DeepSeek's Engram, attach large hashed n-gram tables to a language model, and removing them from DeepSeek-V4.1-Flash costs 36.8 PopQA p…