AI news archive.
19,864 articles, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchACTIVA: Amortized Causal Effect Estimation via Variational Autoencoders
arXiv:2503.01290v4 Announce Type: replace-cross Abstract: Predicting post-intervention distributions from observational data is central to many scientific and decision-making problems, but remains challenging due to caus…
- arXiv cs.AIresearchDoes More Compute Make Every Adversarial Attack Transferable?
arXiv:2504.20310v3 Announce Type: replace-cross Abstract: Transferable attacks, or transferable jailbreaks, are a particularly strong version of adversarial examples that fool all defenders in a class. In this work, we s…
- arXiv cs.AIresearchHow to Backdoor Image Knowledge Distillation
arXiv:2504.21323v4 Announce Type: replace-cross Abstract: Knowledge distillation is widely used to transfer behavior from a large teacher model to a smaller student. It is often assumed to be safe when the teacher is cle…
- arXiv cs.AIresearchDetoxAI: a Python Toolkit for Debiasing Deep Learning Models in Computer Vision
arXiv:2505.05492v2 Announce Type: replace-cross Abstract: While machine learning fairness has made significant progress in recent years, most existing solutions focus on tabular data and are poorly suited for vision-base…
- arXiv cs.AIresearchMAVOS-DD: Multilingual Audio-Video Open-Set Deepfake Detection Benchmark
arXiv:2505.11109v2 Announce Type: replace-cross Abstract: We present the first large-scale open-set benchmark for multilingual audio-video deepfake detection. Our dataset comprises nearly 350 hours of real and fake video…
- arXiv cs.AIresearchShIOEnv: A Command Evaluation Environment for Grammar-Constrained Synthesis and Execution Behavior Modeling
arXiv:2505.18374v3 Announce Type: replace-cross Abstract: We introduce ShIOEnv, a Gymnasium-compatible Bash environment for synthesizing execution-annotated shell input traces. Datasets containing such traces are useful…
- arXiv cs.AIresearchMoVoC: Morphology-Aware Subword Construction for Geez Script Languages Corrected version
arXiv:2509.08812v2 Announce Type: replace-cross Abstract: Subword-based tokenization methods often fail to preserve morphological boundaries, a limitation especially pronounced in low-resource, morphologically complex la…
- arXiv cs.AIresearchInterpreting vision transformers via residual replacement model
arXiv:2509.17401v2 Announce Type: replace-cross Abstract: How do vision transformers (ViTs) represent and process the world? This paper addresses this long-standing question through the first systematic analysis of 6.6K…
- arXiv cs.AIresearchHow Well Does Mathematical Reasoning Transfer Across Languages in English-Centric LLMs?
arXiv:2510.02272v2 Announce Type: replace-cross Abstract: Recent advances in Reinforcement Post-Training (RPT) have substantially improved the mathematical reasoning capabilities of large language models (LLMs), but it r…
- arXiv cs.AIresearchFlowCritic: Bridging Value Estimation with Flow Matching in Reinforcement Learning
arXiv:2510.22686v2 Announce Type: replace-cross Abstract: Reliable value estimation serves as the cornerstone of reinforcement learning (RL) by evaluating long-term returns and guiding policy improvement, significantly i…
- arXiv cs.AIresearchRevisiting GAN with Bayes-Optimal Discrimination
arXiv:2510.25609v4 Announce Type: replace-cross Abstract: We approach generative adversarial network (GAN) training from its original real-versus-fake classification viewpoint, replacing the cross-entropy discriminator o…
- arXiv cs.AIresearchFLAWS: A Benchmark for Error Identification and Localization in Scientific Papers
arXiv:2511.21843v2 Announce Type: replace-cross Abstract: The identification and localization of errors is a core task in peer review, yet the exponential growth of scientific output has made it increasingly difficult fo…
- arXiv cs.AIresearchMarkTune: Improving the Quality-Detectability Trade-off in Model-Embedded LLM Watermarking
arXiv:2512.04044v2 Announce Type: replace-cross Abstract: Language model watermarking schemes fall into two broad categories: inference-time methods, which modify the decoding process at generation time, and model-embedd…
- arXiv cs.AIresearchLLM-Auction: Generative Auction towards LLM-Native Advertising
arXiv:2512.10551v3 Announce Type: replace-cross Abstract: The commercialization of LLM applications is the next frontier in online advertising, with LLM-native advertising emerging as a promising paradigm by integrating…
- arXiv cs.AIresearchCalibrated uncertainty quantification for prosumer flexibility aggregation in ancillary service markets
arXiv:2601.14663v2 Announce Type: replace-cross Abstract: Reliable forecasting of prosumer flexibility is critical for demand response aggregators in frequency controlled ancillary service markets, where reliability requ…
- arXiv cs.AIresearchSpatialMem: Metric-Aligned Long-Horizon Video Memory for Language Grounding and QA
arXiv:2601.14895v4 Announce Type: replace-cross Abstract: SpatialMem is a memory-based system which integrates 3D geometry, semantics, and language with one another in one combined, queryable representation. Starting wit…
- arXiv cs.AIresearchExact and Asymptotically Complete Robust Verifications of Neural Networks via Ising Solvers
arXiv:2603.00408v5 Announce Type: replace-cross Abstract: We present an Ising-compatible framework for formal neural-network robustness verification under bounded input perturbations. For piecewise activations (both cont…
- arXiv cs.AIresearchNear-Optimal Regret for KL-Regularized Multi-Armed Bandits
arXiv:2603.02155v2 Announce Type: replace-cross Abstract: Recent studies have shown that reinforcement learning with KL-regularized objectives can enjoy faster rates of convergence or logarithmic regret, in contrast to t…
- arXiv cs.AIresearchPlace-it-R1: Unlocking Environment-aware Reasoning Potential of MLLM for Video Object Insertion
arXiv:2603.06140v4 Announce Type: replace-cross Abstract: Video object insertion is fundamental to video editing, yet existing diffusion methods often produce visually plausible but physically inconsistent results. We pr…
- arXiv cs.AIagentsOptimistic Policy Regularization: Elite-Statistic Reward Shaping
arXiv:2603.06793v2 Announce Type: replace-cross Abstract: On-policy agents discard trajectories after each update, including rare high-return episodes. We propose Optimistic Policy Regularization (OPR), which turns those…
- arXiv cs.AIresearchMarkovian Generation Chains in Large Language Models
arXiv:2603.11228v2 Announce Type: replace-cross Abstract: The widespread use of large language models (LLMs) raises an important question: how do texts evolve when they are repeatedly processed by LLMs? In this paper, we…
- arXiv cs.AIresearchA Sociolinguistic Analysis of Automatic Speech Recognition Bias in Newcastle English
arXiv:2603.24549v2 Announce Type: replace-cross Abstract: Automatic Speech Recognition (ASR) systems are widely used in everyday communication, education, healthcare, and industry, yet their performance remains uneven ac…
- arXiv cs.AIagentsMeasuring the Permission Gate: A Stress-Test Evaluation of Claude Code's Auto Mode
arXiv:2604.04978v3 Announce Type: replace-cross Abstract: Claude Code's auto mode is the first deployed permission system for AI coding agents, using a two-stage transcript classifier to gate dangerous tool calls. Anthro…
- arXiv cs.AIresearchAtManRL: Towards Faithful Reasoning via Differentiable Attention Saliency
arXiv:2604.16158v2 Announce Type: replace-cross Abstract: Large language models (LLMs) increasingly rely on chain-of-thought (CoT) reasoning to solve complex tasks, yet a reasoning trace may merely accompany the final an…
- arXiv cs.AIresearchGraph Memory Transformer (GMT)
arXiv:2604.23862v4 Announce Type: replace-cross Abstract: We investigate whether an explicit learned memory graph can replace the Feed-Forward Network (FFN) sublayer in a decoder-only transformer while preserving the sur…
- arXiv cs.AIresearchUnifying Goal-Conditioned RL and Unsupervised Skill Learning via Control-Maximization
arXiv:2605.06145v2 Announce Type: replace-cross Abstract: Unsupervised pretraining has driven empirical advances in goal-conditioned reinforcement learning (GCRL), but its theoretical foundations remain poorly understood…
- arXiv cs.AIresearchBenchmarking Compositional Generalisation for Machine Learning Interatomic Potentials
arXiv:2605.08988v2 Announce Type: replace-cross Abstract: Machine Learning Interatomic Potentials play a fundamental role in computational chemistry and materials science, enabling applications from molecular dynamics si…
- arXiv cs.AIresearchDisentangled Sparse Representations for Concept-Separated Diffusion Unlearning
arXiv:2605.12122v3 Announce Type: replace-cross Abstract: Unlearning specific concepts in text-to-image diffusion models has become increasingly important for preventing undesirable content generation. Among prior approa…
- arXiv cs.AIresearchManifold Sampling via Entropy Maximization
arXiv:2605.12338v2 Announce Type: replace-cross Abstract: Sampling from constrained distributions has a wide range of applications, including in Bayesian optimization and robotics. Prior work establishes convergence and…
- arXiv cs.AIresearchLearning Unbiased Permutations via Flow Matching
arXiv:2605.16755v2 Announce Type: replace-cross Abstract: Learning permutations is fundamental to sorting, ranking, and matching, but existing differentiable methods based on entropy-regularized Sinkhorn produce a single…
- arXiv cs.AIagentsStop Drawing Scientific Claims from LLM Social Simulations Without Robustness Audits
arXiv:2605.18890v2 Announce Type: replace-cross Abstract: The scientific claims drawn from LLM social simulations should be no stronger than the robustness audits that support them. Generative agents bring new expressive…
- arXiv cs.AIresearchFormalASR: End-to-End Spoken Chinese to Formal Text
arXiv:2605.19266v4 Announce Type: replace-cross Abstract: Automatic speech recognition (ASR) systems are typically optimized for verbatim transcription, which preserves disfluencies, filler words, and informal spoken str…
- arXiv cs.AIresearchWhen Should LLMs Reason? Adaptive Inference Routing via Entropy Dynamics
arXiv:2605.22873v3 Announce Type: replace-cross Abstract: Chain-of-thought (CoT) reasoning can improve LLM performance, but its benefits vary across tasks, models, and individual queries. On some tasks, CoT provides marg…
- arXiv cs.AIagentsHow Far Will They Go? Red-Teaming Online Influence with Large Language Models
arXiv:2605.22880v2 Announce Type: replace-cross Abstract: As large language model (LLM)-based agents increasingly participate in online discourse, red-teaming their capacity to support political influence campaigns is cr…
- arXiv cs.AIresearchConfidence and Calibration of Activation Oracles for Reliable Interpretation of Language Model Internals
arXiv:2605.26045v3 Announce Type: replace-cross Abstract: An activation oracle is a language model trained to convert the activations of another model into natural language. However, it is so far unclear to what extent t…
- arXiv cs.AIresearchSemantic Optimal Transport for Sparse Autoencoder Feature Matching and Circuit Compression
arXiv:2605.28567v3 Announce Type: replace-cross Abstract: Sparse autoencoders (SAEs) have become a central tool for interpreting language models. However, two key SAE analyses that remain difficult to scale are (1) match…
- arXiv cs.AIresearchRethinking Molecular Text Representations for LLMs: An Empirical Study
arXiv:2606.03057v2 Announce Type: replace-cross Abstract: Large language models (LLMs) are increasingly used for molecular tasks, but it remains unclear which molecular representation to use. We present a systematic benc…
- arXiv cs.AIresearchBehavioral Consistency in Machine Learning
arXiv:2606.07581v2 Announce Type: replace-cross Abstract: A trained model is usually described by its weights, as if the weights fixed its behavior. In practice the same weights are run by different programs. Training us…
- arXiv cs.AIresearchHacking Generative Perplexity: Why Unconditional Text Evaluation Needs New Metrics
arXiv:2606.08417v3 Announce Type: replace-cross Abstract: Diffusion language models (DLMs) have emerged as the leading non-autoregressive alternative to language modeling. Because their likelihoods are intractable, gener…
- arXiv cs.AIresearchMeasuring Pretraining Contamination in Public Medical Vision-Language Benchmarks
arXiv:2606.10066v2 Announce Type: replace-cross Abstract: Medical vision-language models (VLMs) are evaluated on public benchmarks whose images and question-answer pairs have been freely available for years, yet reported…
- arXiv cs.AIresearchBeyond Prompt and Parser Engineering: Soft-Prompt Tuning for Automated LLM Evaluation
arXiv:2606.12117v2 Announce Type: replace-cross Abstract: Benchmark scores often misrepresent a large language model's (LLM's) knowledge, because they rely, e.g., on the model's ability to follow specific formatting requ…
- arXiv cs.AIagentsPerceive, Interact, Reason: Building Tool-Augmented Visual Agents for Spatial Reasoning
arXiv:2606.12830v2 Announce Type: replace-cross Abstract: While recent vision-language models (VLMs) demonstrate strong multimodal understanding, they remain limited in spatial reasoning tasks that require active evidenc…
- arXiv cs.AIresearchSDS-LoRA: Overcoming Anisotropic Gradient Scaling in Low-Rank Adaptation
arXiv:2606.16454v3 Announce Type: replace-cross Abstract: Low-Rank Adaptation (LoRA) enables efficient adaptation of large pretrained models to downstream tasks by parameterizing weight updates with low-rank matrices. In…
- arXiv cs.AIresearchDescriptor: Certus Caliber Classification Gunshot Dataset (C3GD)
arXiv:2606.18135v2 Announce Type: replace-cross Abstract: In this work, we introduce the Certus Caliber Classification Gunshot Dataset (C3GD), a publicly accessible dataset developed for the analysis of firearm muzzle bl…
- arXiv cs.AIresearchMotionHalluc: Diagnosing Kinematic Hallucinations in Fine-Grained Motion Reasoning
arXiv:2606.23061v2 Announce Type: replace-cross Abstract: Motion instruction generation in cross-video comparison aims to produce corrective feedback that describes the differences between a query and a reference motion.…
- arXiv cs.AIresearchReactiveBFM: Reactive Closed-Loop Motion Planning Towards Universal Humanoid Whole-Body Control
arXiv:2606.30362v3 Announce Type: replace-cross Abstract: While current Behavior Foundation Models (BFMs) provide robust control priors for humanoids, they only execute pre-defined reference motions. As a result, they ar…
- arXiv cs.AIresearchPosterHarness: Turning Scientific Poster Generation into an Auditable Instruction-Following Benchmark
arXiv:2607.03006v2 Announce Type: replace-cross Abstract: Text-rich image models can now design poster-scale layouts, but we lack ways to measure whether they honor scientific communication contracts: legible labels, pre…
- arXiv cs.AIagentsRobustness Verification of an Autonomous Underwater Vehicle-based Plankton Classifier
arXiv:2607.04453v2 Announce Type: replace-cross Abstract: The assessment of planktonic standing stocks and microorganism structures is critical for understanding upper ocean biological processes. Currently, autonomous un…
- arXiv cs.AIresearchContext-Aware Subtitle Translation on CPU: Deployment-Driven Training, Evaluation, and Latency Design
arXiv:2607.09957v3 Announce Type: replace-cross Abstract: We study English-to-Traditional-Chinese (Taiwan) subtitle translation for a free browser extension, where the model must run on the viewer's CPU in about 500 ms p…
- arXiv cs.AIagentsARES: Adaptive Reasoning-Effort Steering for PPA- and Cost-Aware RTL Optimization with LLM Agents
arXiv:2607.27879v2 Announce Type: replace-cross Abstract: Large language model (LLM) agents optimize the power, performance, and area (PPA) of register-transfer-level (RTL) designs by iterating over edits, synthesis, and…