AI news archive.
10,659 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchIAD-Unify: Task-Specific Interfaces for Industrial Anomaly Understanding, Segmentation, and Generation
arXiv:2604.12440v3 Announce Type: replace-cross Abstract: Industrial anomaly inspection requires complementary capabilities: explaining an observed defect, localizing its pixels, and synthesizing a controlled edit. We pr…
- arXiv cs.AIresearchSparKV: Overhead-Aware KV Cache Loading for Efficient On-Device LLM Inference
arXiv:2604.21231v3 Announce Type: replace-cross Abstract: Efficient inference for on-device Large Language Models (LLMs) remains challenging due to limited hardware resources and the high cost of the prefill stage, which…
- arXiv cs.AIresearchSAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild
arXiv:2605.07604v2 Announce Type: replace-cross Abstract: 3D animal reconstruction in the wild remains challenging due to large species variation, frequent occlusions, and the prevalence of multi-animal scenes, while exi…
- arXiv cs.AIresearchAPEX: Assumption-free Projection-based Embedding eXamination Metric for Generative Distribution Evaluation
arXiv:2605.07786v4 Announce Type: replace-cross Abstract: As generative models achieve unprecedented visual quality, the gold standard for image evaluation remains traditional feature-distribution metrics. However, these…
- arXiv cs.AIresearchDCVD: Dual-Channel Cross-Modal Fusion for Joint Vulnerability Detection and Localization
arXiv:2605.11015v2 Announce Type: replace-cross Abstract: Software vulnerability detection plays a critical role in ensuring system security, where real-world auditing requires not only determining whether a function is…
- arXiv cs.AIresearchFeatCal: Feature Calibration for Post-Merging Models
arXiv:2605.13030v2 Announce Type: replace-cross Abstract: Model merging combines task experts into one model and avoids joint training, retraining, or deploying many expert models, but the merged model often still underp…
- arXiv cs.AIresearchBeyond Accuracy: Robustness, Interpretability and Expressiveness of EEG Foundation Models
arXiv:2605.17562v2 Announce Type: replace-cross Abstract: EEG foundation models (EEG-FMs) have been evaluated predominantly on clean, in-distribution accuracy, demonstrating modest gains over supervised baselines and wea…
- arXiv cs.AIresearchSelf-Evolving Spatial Reasoning in Vision Language Models via Geometric Logic Consistency
arXiv:2605.18162v2 Announce Type: replace-cross Abstract: Vision-Language Models (VLMs) have made striking progress, yet their spatial reasoning remains fragile. Models that answer an original input correctly can still f…
- arXiv cs.AIresearchToken by Token, Compromised: Backdoor Vulnerabilities in Unified Autoregressive Models
arXiv:2605.19227v2 Announce Type: replace-cross Abstract: Unified autoregressive models (UAMs) are transformer models that generate text as well as image tokens within a single autoregressive pass. Shared parameters and…
- arXiv cs.AIresearchWorld-Ego Modeling for Embodied Video Generation in Long-Horizon Navigation-Manipulation Tasks
arXiv:2605.19957v2 Announce Type: replace-cross Abstract: Embodied video world models typically capture both scene evolution and the robot's behavior, which we refer to as the \emph{world} and the \emph{ego}, respectivel…
- arXiv cs.AIresearch4D-GSW: Kinematic-Aware Spatio-Temporal Consistent Watermarking for 4D Gaussian Splatting
arXiv:2605.22342v2 Announce Type: replace-cross Abstract: While 4D Gaussian Splatting (4DGS) has revolutionized high-fidelity dynamic reconstruction, safeguarding the intellectual property of these assets remains an open…
- arXiv cs.AIresearchThe Distillation Game: Adaptive Evaluations & Efficient Defenses
arXiv:2605.22737v4 Announce Type: replace-cross Abstract: Distillation attacks create a deployment trade-off for model providers: the same outputs that make a model more useful can also make it easier to imitate. We stud…
- arXiv cs.AIresearchROAR: Retrieval Opportunity-Aware Refinement for Zero-Shot Time Series Forecasting
arXiv:2605.24911v2 Announce Type: replace-cross Abstract: Retrieval augmentation provides time series forecasters with historical continuations, yet even candidates that outperform the base forecast may fail to improve t…
- arXiv cs.AIresearchTrust Region Q Adjoint Matching
arXiv:2605.27079v2 Announce Type: replace-cross Abstract: Off-policy reinforcement learning of pretrained flow policies remains challenging due to the instability of optimization arising from the multi-step sampling proc…
- arXiv cs.AIresearchEfficient and Scalable Provenance Tracking for LLM-Generated Code Snippets
arXiv:2605.28510v3 Announce Type: replace-cross Abstract: Large language models (LLMs) for code completion and generation are increasingly used in software development, yet they may reproduce training examples verbatim a…
- arXiv cs.AIresearchMemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems
arXiv:2605.28732v4 Announce Type: replace-cross Abstract: Memory is essential for enabling large language models to support long-horizon reasoning, yet existing memory systems remain unreliable and difficult to debug. Tr…
- arXiv cs.AIresearchLearning Implicit Bias in Generative Spaces for Accelerating Protein Dynamics Emulation
arXiv:2606.01833v2 Announce Type: replace-cross Abstract: Generative emulators of protein dynamics produce plausible trajectories at a fraction of the cost of molecular dynamics, but they inherit their training distribut…
- arXiv cs.AIresearchLoRi: Low-Rank Distillation for Implicit Reasoning
arXiv:2606.05315v2 Announce Type: replace-cross Abstract: Implicit chain-of-thought (iCoT) methods aim to internalize reasoning in large language models, but often underperform explicit CoT prompting. We empirically find…
- arXiv cs.AIresearchLearning to Attack and Defend: Adaptive Red Teaming of Language Models via GRPO
arXiv:2606.09701v2 Announce Type: replace-cross Abstract: Language model safety must continually adapt to evolving attacks. Recent works have demonstrated that reinforcement learning can be used to train stronger attacke…
- arXiv cs.AIresearchPretrained self-supervised speech models can recognize unseen consonants
arXiv:2606.11542v2 Announce Type: replace-cross Abstract: Modern pretrained self-supervised automatic speech recognition models are trained on large-scale audio data to encode speech into contextualized representations.…
- arXiv cs.AIresearchAnalyzing and Improving Fine-grained Preference Optimization in Medical LVLMs
arXiv:2606.12590v2 Announce Type: replace-cross Abstract: Preference optimization is increasingly used to post-train medical large vision-language models (LVLMs), yet it operates at a much coarser granularity than the on…
- arXiv cs.AIresearchLearned Image Compression for Vision-Language-Action Models
arXiv:2606.16253v2 Announce Type: replace-cross Abstract: Vision-language-action (VLA) models increasingly rely on high-frequency multi-camera observations, making visual communication a major bottleneck for real-time ro…
- arXiv cs.AIresearchPearlVLA: Progressive Embodied Action-Plan Refinement in Latent Space
arXiv:2606.17924v2 Announce Type: replace-cross Abstract: Current Vision-Language-Action (VLA) models face a trade-off between efficient action generation and explicit deliberation. Directly decoding actions from vision-…
- arXiv cs.AIresearchS4oP: Operator-level Pruning of Structured State Space Models for Resource-Constrained Devices
arXiv:2606.18096v2 Announce Type: replace-cross Abstract: Structured State Space Models (SSMs), including the S4 and S4D architectures, have recently emerged as powerful alternatives to attention-based models for capturi…
- arXiv cs.AIresearchLearning from Own Solutions: Self-Conditioned Credit Assignment for Reinforcement Learning with Verifiable Rewards
arXiv:2606.18810v2 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has driven substantial progress in training LLMs for reasoning tasks, but representative methods such as GRP…
- arXiv cs.AIresearchNeural Conjugate Aggregation: Identifiable Unsupervised Multi-Sensor Regression under Heterogeneous Sensor Bias
arXiv:2606.22200v2 Announce Type: replace-cross Abstract: We study regression-based data fusion under uncertainty, where multiple noisy and biased measurement sources are available but ground-truth labels are absent duri…
- arXiv cs.AIresearchReward Observability and the Limits of Offline Checkpoint Selection in RSSM World Models
arXiv:2607.01736v3 Announce Type: replace-cross Abstract: We study the closed-loop properties of a recurrent state-space model (RSSM) world model trained on human demonstrations in Gymnasium's LunarLander-v3. We use the…
- arXiv cs.AIresearchACID: Action Consistency via Inverse Dynamics for Planning with World Models
arXiv:2607.02403v2 Announce Type: replace-cross Abstract: Decision-time planning with action-conditioned world models has become a popular paradigm for embodied control. However, the standard planning cost judges a candi…
- arXiv cs.AIresearchBeyond Euclidean Clipping: Overcoming Exploration Collapse in LLM RL via Riemannian Isometric Policy Optimization
arXiv:2607.10169v2 Announce Type: replace-cross Abstract: Reinforcement learning (RL) has become a dominant paradigm for enhancing LLMs' reasoning capabilities. However, RL algorithms with PPO-Clip are inherently limited…
- arXiv cs.AIresearchIntroducing Human-Centeredness in AI-Assisted Lexicography
arXiv:2607.11808v4 Announce Type: replace-cross Abstract: This paper proposes a human-centered artificial intelligence (HCAI) framework for AI-assisted lexicography. While generative AI offers significant opportunities t…
- arXiv cs.AIresearchThe One-Word Census: Answer-Choice Conformity Across 105 Language Models
arXiv:2607.12796v4 Announce Type: replace-cross Abstract: When a language model must choose one answer from a large space of equally valid options, which answer does it choose, and how often is it the answer every other…
- arXiv cs.AIresearchWhat Transfers from Text to Vision? Capability Scaling Laws and Transfer Dynamics for VLMs
arXiv:2608.00013v4 Announce Type: replace-cross Abstract: Choosing the right large language model (LLM) backbone is the most consequential decision when building a vision-language model (VLM), yet it remains fundamentall…
- arXiv cs.AIresearchRobust and Efficient Noisy-Label Time-Series Classification via Dynamic Time Warping Based Granular Ball Computing
arXiv:2608.11704v2 Announce Type: replace-cross Abstract: Dynamic Time Warping (DTW)-based Nearest-Neighbor (NN) classifiers are effective for time-series classification but are vulnerable to mislabeled training samples…
- arXiv cs.AIresearchCoToGrasp: Contact-Topology-Conditioned Dexterous Grasp Synthesis via Canonical Workspace Learning
arXiv:2608.19776v3 Announce Type: replace-cross Abstract: Current dexterous grasp planners primarily optimize for physical stability, focusing on whether an object can be grasped rather than how it should be grasped to s…
- arXiv cs.AIresearchA Network Science Perspective on Evaluating Deep Graph Generative Models
arXiv:2609.01015v2 Announce Type: replace-cross Abstract: Traditional network models from network science, such as the Erdos-Renyi and configuration models, generate random networks that reproduce few selected topologica…
- arXiv cs.AIresearchMoEMB: Scaling Universal Multimodal Embeddings with Efficient Mixture-of-Experts Models
arXiv:2609.08663v2 Announce Type: replace-cross Abstract: Universal multimodal embedding (UME) increasingly demands encoder's capacity for handling a broad range of tasks and modalities with increased complexity. Prior s…
- arXiv cs.AIresearchThe Truth Was Never Gone: Perfect Aliasing in Compliant-Context Truth Probes
arXiv:2609.10739v2 Announce Type: replace-cross Abstract: Linear probes that decode the truth from a language model's activations have been proposed as deception monitors, and a truth probe scoring below chance on a mode…
- arXiv cs.AIresearchMeddies-PII: A Multilingual Framework for Personally Identifiable Information Extraction in Clinical De-identification
arXiv:2609.12544v2 Announce Type: replace-cross Abstract: Clinical de-identification relies on accurately identifying personally identifiable information (PII). However, manually annotated datasets are costly to construc…
- LessWrongresearchThree Pieces of Evidence that the NLA Reconstructor Does Not Read Semantics or Syntax
TL;DR The autoreconstructor (AR) in a natural language autoencoder (NLA) is the entire argument for the validity of the method, and at present is poorly understood. The AR does not condition strongly on English grammar o…
- MIT Technology Review AIresearchRoundtables: A Conversation With the Creator of AI-Designed Viruses
Friday, October 16, 2026 Can AI design new life forms? In 2025, Stanford University PhD student Samuel King came up with a preliminary answer when he used a generative AI model to propose genetic blueprints for microscop…
- LessWrongresearchSocial Theory and Agent Foundations
Lately I’ve been writing a lot about the sociopolitical considerations that economic-style reasoning misses . I view this work as continuous with my agent foundations research; in this post, I lay out some of the connect…
- Engadget AIresearchFired OpenAI safety researchers dispute their dismissals in open letter
Three OpenAI researchers said their firings may have a 'chilling' effect on employees.
- LessWrongresearchReviewing failure modes in science as we prepare for AI researchers
Summary : Science has a long list of documented failure modes, such as p-hacking, the File Drawer Problem, shallow peer review, and ignored retractions. This post catalogs about 50 of them with the evidence for each, the…
- LessWrongresearch[Paper] J++ Lens: Jacobian Filtering Enables More Faithful Workspace Lenses
Kola Ayonrinde, Anthropic Fellows, koayon@gmail.com ; Jack Lindsey, Anthropic TL;DR : The J-Lens was proposed to read verbalisable representations from language model activations. We introduce the J++ Lens: an improvemen…
- LessWrongresearchJ++ Lens: Jacobian Filtering Enables More Faithful Workspace Lenses
Kola Ayonrinde, Anthropic Fellows, koayon@gmail.com ; Jack Lindsey, Anthropic TL;DR : The J-Lens was proposed to read verbalisable representations from language model activations. We introduce the J++ Lens: an improvemen…
- LessWrongresearch5-MeO-DMT sometimes cancels the Ebbinghaus illusion
Psychedelics produce all sorts of visual distortions: from breathing walls and geometric patterns to complex visual landscapes populated by full-blown entities. Sometimes people claim to see the Truth on them — but it’s…
- The DecoderresearchSome mathematicians call for OpenAI boycott after AI-generated proofs flood their field
The Association for Human Mathematics is calling for an OpenAI boycott after the company released more than 700 AI-generated math manuscripts at once. OpenAI had to retract three papers the next day over a sign error. Fi…
- LessWrongresearchYou Heard of "Bayesians Don't Predictably Update" Because of Robin Hanson
Robin Hanson didn't invent the idea of a martingale process. The lilim say that Ville in 1939 gave the term its mathematical meaning and Doob turned it into the modern theory. But to state this is not to realize that you…
- TechCrunch AIresearchOpenAI’s math solutions aren’t meeting the field’s standards yet
OpenAI's flood of proofs deviated from the guidelines set by a group of mathematical researchers consulted by the frontier lab.
- Latent SpaceresearchSynthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
A special Science pod and Engineering pod crossover.. with Forward Deployed Engineering kicker!