AI news archive.
10,659 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchBeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series
arXiv:2610.05409v1 Announce Type: cross Abstract: Action-conditioned time-series forecasting requires accounting for how future actions and exogenous forcings influence multiple targets through partially observed effects…
- arXiv cs.AIresearchA Strong Baseline for Evaluating Vision Encoders in Multimodal Large Language Models
arXiv:2610.05413v1 Announce Type: cross Abstract: Evaluating vision encoders requires metrics that reliably predict their downstream performance in multimodal large language models (MLLMs). Although recent studies have s…
- arXiv cs.AIresearchHierarchical Time-aware Bootstrapping for Off-Policy Subgoal Value Learning
arXiv:2610.05446v1 Announce Type: cross Abstract: Off-policy hierarchical reinforcement learning must estimate the values of high-level decisions while the low-level policy changes. HIRO adapts replay data through subgoa…
- arXiv cs.AIresearchHuman-Like Attention? A Psychophysical Comparison of Visual Search in Humans and MLLMs
arXiv:2610.05463v1 Announce Type: cross Abstract: Visual search is a fundamental cognitive ability. This study investigates whether Multimodal Large Language Models (MLLMs) exhibit human-like difficulty signatures in vis…
- arXiv cs.AIresearchPopulation Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning
arXiv:2610.05476v2 Announce Type: replace-cross Abstract: Scaling decentralized learning changes not only the number of clients $N$, but also the dynamics of information propagation and consensus. We argue that the effec…
- arXiv cs.AIresearchReflecting on Creative-Boundaries with an AI Co-Doodler
arXiv:2610.05482v1 Announce Type: cross Abstract: In this pictorial, we consider how the negotiation of creative boundaries with a co-creative AI system can create moments for personal creative reflection. We ground this…
- arXiv cs.AIresearchFLEX-WAM: Flexible Block-Causal World-Action Models for Long-Horizon Imagination and Planning
arXiv:2610.05483v1 Announce Type: cross Abstract: World--action models (WAMs) promise a unified model that predicts action-conditioned futures, generates feasible actions, and supports planning in imagination. However, e…
- arXiv cs.AIresearchWhat Does a Harness Repair? Three Preregistered Studies Around a Longevity-Tuned Language Model
arXiv:2610.05533v3 Announce Type: replace-cross Abstract: A harness search around a frozen longevity-tuned 1.7B model reported a rescue. An audit found that the gain came from rows the start left unparsed, that the calle…
- arXiv cs.AIresearchLiFT: Loop Flow Transformers
arXiv:2610.05538v1 Announce Type: cross Abstract: We introduce Loop Flow Transformers (LiFT), a family of looped generative models that scales computation by repeatedly applying a shared Diffusion Transformer (DiT) core,…
- arXiv cs.AIresearchDon't Judge an LLM Only by Its Activations: Discovering Suppressed Safety Features via Counterfactual Activation Potential
arXiv:2610.05541v1 Announce Type: cross Abstract: Mechanistic interpretability has emerged as the primary means to understand safety behavior of LLMs. However, existing tools primarily focus on the activating neurons or…
- arXiv cs.AIresearchSoft Strategy Selection for Batch-Mode Active Learning
arXiv:2610.05544v1 Announce Type: cross Abstract: Real-world deployment of active learning typically forces practitioners to choose an acquisition strategy before any data is labeled. This is a daunting task: strategy pe…
- arXiv cs.AIresearchCut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation
arXiv:2610.05559v1 Announce Type: cross Abstract: Industrial sequential recommender systems operate over massive item catalogs (e.g., 10^5--10^7 items). Multi-label recommendation models are trained with Binary Cross-Ent…
- arXiv cs.AIresearchCan Prosodic Style Be Inferred from Text Alone? Evidence from Unsupervised Acoustic Clusters
arXiv:2610.05575v1 Announce Type: cross Abstract: Much of expressive text-to-speech research rests on an untested assumption that written text carries enough information to select an appropriate prosodic style for its de…
- arXiv cs.AIresearchWhat Does an Observability Foundation Model Know?
arXiv:2610.05577v1 Announce Type: cross Abstract: A linear probe can show that a label is recoverable from a model's hidden states, but not whether that goes beyond what the input already reveals, or whether the model us…
- arXiv cs.AIresearchExpanding LLM Reasoning
arXiv:2610.05584v1 Announce Type: cross Abstract: Extra inference compute is usually spent on sampling more reasoning chains. We study where inside an existing chain an additional continuation should begin. We define exp…
- arXiv cs.AIresearchWhen the Cross-Silo Federation Goes Offline: Continual Learning for Site Onboarding with Limited Unlabeled Data
arXiv:2610.05598v1 Announce Type: cross Abstract: An organization often holds too little labeled data to train a model that generalizes, and the records that would supply the rest sit with organizations that cannot relea…
- arXiv cs.AIresearchAn LLM-in-the-loop RL Framework for Bioinformatics Feature Selection
arXiv:2610.05600v1 Announce Type: cross Abstract: High-dimensional bioinformatics data, characterized by a large number of features relative to the number of samples, pose major challenges such as the ``curse of dimensio…
- arXiv cs.AIresearchYour Unlearning Gives You Away: Identifying Erased Concepts in Diffusion Models
arXiv:2610.05601v1 Announce Type: cross Abstract: Existing attacks on unlearned diffusion models assume that the erased concepts are known in advance and focus on recovering them. In practice, however, model providers ma…
- arXiv cs.AIresearchKandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation
arXiv:2610.05608v1 Announce Type: cross Abstract: We present Kandinsky 6.0 Video, a family of foundation diffusion models for synchronized text-to-audio-video generation, comprising Kandinsky 6.0 Video Lite (3B parameter…
- arXiv cs.AIresearchSEA-LM: Egocentric Spatial Audio Understanding for Wearable Microphone Arrays
arXiv:2610.05610v2 Announce Type: replace-cross Abstract: Embodied, ego-centric intelligence fundamentally requires the ability to comprehend spatial audio within complex environments. While large audio-language models e…
- arXiv cs.AIresearchTransformer-Based Time-Series Inference of Lindblad Dynamics in Open Quantum Systems
arXiv:2610.05647v1 Announce Type: cross Abstract: The Lindblad master equation is the standard framework for describing the non-unitary evolution of open quantum systems, where environmental interactions induce dissipati…
- arXiv cs.AIresearchKnowing the Rules, Applying the Rules: Evaluating Language Models on Traditional Chinese Bazi
arXiv:2610.05682v1 Announce Type: cross Abstract: Knowing domain rules does not guarantee applying them to a case. We study this distinction in traditional Chinese Bazi through 3,000 Chinese multiple-choice questions spa…
- arXiv cs.AIresearchTwo-Sample Testing via Path-based Inference
arXiv:2610.05684v1 Announce Type: cross Abstract: Modern deep generative models are primarily studied for their ability to generate realistic samples, yet the generative dynamics they learn can also serve as objects of s…
- arXiv cs.AIresearchSimpleMark: Fast Multi-Bit Text Watermarking under f -Divergence Constraints
arXiv:2610.05712v1 Announce Type: cross Abstract: We introduce a framework for multi-bit text watermarking with security defined directly through $f$-divergence from the base language model distribution. Unlike prior app…
- arXiv cs.AIresearchRotated, but How Far? Diagnosing and Improving Object-Rotation Reasoning in VLMs
arXiv:2610.05715v1 Announce Type: cross Abstract: Vision-language models (VLMs) can detect that an object has rotated across views, but cannot reliably tell by how much. We introduce OR-Bench, a fine-grained benchmark fo…
- arXiv cs.AIresearchHLA-WM: Hybrid Linear Attention for Long-Horizon Video World Models
arXiv:2610.05739v1 Announce Type: cross Abstract: Long-horizon video world models require persistent memory to preserve scene consistency over extended rollouts. Softmax attention retains the full generation history thro…
- arXiv cs.AIresearchCIPHER-MoE: Balancing Efficiency and Routing Fidelity in Trillion-Scale MoE Training
arXiv:2610.05744v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) has been widely adopted in recent large language model (LLM) architectures. However, scaling up MoE in LLM training introduces system-level chall…
- arXiv cs.AIresearchRelational Synthesis: Structure-Mediated Concatenative Synthesis for Foley and Retrieval-Augmented Audio Generation
arXiv:2610.05768v1 Announce Type: cross Abstract: We ask: given a retrieved source audio $S$ and a separate reference audio $R$, can we synthesize novel audio $Y$ out of this pair $(S,R)$ such that $Y$ remains acoustical…
- arXiv cs.AIresearchCLARA: Can AI Assess Developmental Appropriateness in Children's Stories?
arXiv:2610.05783v1 Announce Type: cross Abstract: Assessing the developmental suitability of children's narratives is important for educational recommendation and developmental literacy research, yet such assessment typi…
- arXiv cs.AIresearchProtocol-Sensitive Evaluation of Log Anomaly Detection: Component Costs and Target-Access Sensitivity on HDFS and BGL
arXiv:2610.05807v1 Announce Type: cross Abstract: Protocol choices can change the conclusions drawn from log anomaly detection benchmarks even when detector settings are fixed. We present a joint empirical study of split…
- arXiv cs.AIresearchLevel-of-Token Diffusion
arXiv:2610.05816v1 Announce Type: cross Abstract: Image and video diffusion models allocate equal computation to every region, even when the intended scene calls for varying levels of detail. The spatial distribution of…
- arXiv cs.AIresearchCACFG: Curvature-Aware Classifier-Free Guidance and Optimal Control
arXiv:2610.05845v1 Announce Type: cross Abstract: Diffusion models generate samples by learning to reverse a fixed corruption process, and classifier-free guidance (CFG) is the standard mechanism for conditioning this pr…
- arXiv cs.AIresearchHierarchical Reinforcement Learning for Collision-Free Locomotion of an Underactuated Biped
arXiv:2610.05855v1 Announce Type: cross Abstract: A bipedal robot cannot deviate from its path to avoid an obstacle without disturbing its balance, and this coupling is most severe on underactuated platforms such as the…
- arXiv cs.AIresearchCurriculum Brain: Constructing Curriculum Knowledge Graphs as a Substrate for Cognitive Diagnosis
arXiv:2610.05860v1 Announce Type: cross Abstract: Cognitive Diagnostic Models (CDMs) identify which specific skills a student has and has not mastered, the signal a personalized learning path needs and a single aggregate…
- arXiv cs.AIresearchCCQ: A Multi-State Child Care Quality Dataset to Support AI for Children's Health Research
arXiv:2610.05863v1 Announce Type: cross Abstract: High-quality child care in early life is a critical determinant of children's growth and development. Research on child care quality has been constrained by fragmented, n…
- arXiv cs.AIresearchWeave Mamba Fusion: Global Cross-Scale Interaction for Lightweight Face Detection
arXiv:2610.05865v1 Announce Type: cross Abstract: Feature pyramid methods, from FPN to BiFPN, have achieved strong performance in face detection by fusing multi-scale features. However, detecting faces under unconstraine…
- arXiv cs.AIresearchOff-Policy Merging Beats On-Policy Self-Distillation for Continual Learning
arXiv:2610.05872v1 Announce Type: cross Abstract: A long-standing goal of AI is a model that can continually learn and improve itself. On post-trained models, supervised finetuning (SFT) on new data often causes poor gen…
- arXiv cs.AIresearchOGAM: Connecting Systematic Testing to Runtime Assurance through Object-Grounded Attention Monitoring for VLA Policies
arXiv:2610.05878v1 Announce Type: cross Abstract: Benchmarks expose vision-language-action (VLA) policies to few canonical instructions, while exhaustive deployment testing is impossible. We introduce Object-Grounded Att…
- arXiv cs.AIresearchTasteRoute: Personalized Routing for Video Generation
arXiv:2610.05896v1 Announce Type: cross Abstract: Rapid progress in video generation has led to a plethora of models that differ substantially in capability and generation cost. This raises a natural question: can each r…
- arXiv cs.AIresearchOn Hyperparameter Tuning on the Test Set
arXiv:2610.05902v1 Announce Type: cross Abstract: "Don't tune hyperparameters on the test set" is often stated in machine learning textbooks. Violating it is considered a cardinal sin that produces misleadingly optimisti…
- arXiv cs.AIresearchNon-invasive Seizure Detection Using Wearable Wrist-worn Accelerometry and Deep Learning
arXiv:2610.05919v1 Announce Type: cross Abstract: Seizure monitoring and detection are crucial for reducing the morbidity and mortality associated with seizures. Current epilepsy care, often involving expensive video-ele…
- arXiv cs.AIresearchIncentive Alignment in Online Experimentation
arXiv:2610.05922v2 Announce Type: replace-cross Abstract: Evaluating the causal effect of new features is a central goal for online platforms. While recent literature addresses limited testing traffic via centralized por…
- arXiv cs.AIresearchUltraDub: Towards Authentic Dubbing by Unifying Visually-Steered Flow Learning and Trajectory Guidance
arXiv:2610.05932v1 Announce Type: cross Abstract: Visual voice cloning requires intelligible, speaker-consistent speech synchronized with visible articulation. However, sequential multimodal conditioning can disrupt prev…
- arXiv cs.AIresearchReMem: Streaming Video Understanding With Long Context Retention
arXiv:2610.05940v1 Announce Type: cross Abstract: Despite their impressive performance on a wide range of video understanding tasks, current Vision Language Models (VLMs) are predominantly designed for offline scenarios…
- arXiv cs.AIresearchPhysics-Informed but Not Physics-Consistent: Error Geometry and Subspace Projection for Neural AC Power Flow
arXiv:2610.05959v1 Announce Type: cross Abstract: Recent neural power-flow solvers, including emerging foundation models, achieve accurate voltage predictions, yet such accuracy does not necessarily imply physically cons…
- arXiv cs.AIresearchHuatuoGPT-3: RL-Only Domain Adaptation from Base Models
arXiv:2610.05966v1 Announce Type: cross Abstract: Domain adaptation aims to turn a general-purpose large language model (LLM) into an expert for a target domain. While the dominant SFT+RL pipeline offers a convenient col…
- arXiv cs.AIresearchJLD: Perceptual Distance Through A Jacobian Lens
arXiv:2610.05967v1 Announce Type: cross Abstract: Image compression, restoration, and generation all require a way to measure how different two images look to a person. Pixel error ignores how people see, while the most…
- arXiv cs.AIresearchTransfer-Stratified On-Policy Distillation for RL-Improved Reasoning Teachers
arXiv:2610.05974v1 Announce Type: cross Abstract: Reinforcement learning can substantially improve a reasoning teacher, but it is unclear which of those improvements survive when the teacher supervises a smaller on-polic…
- arXiv cs.AIresearchEpicWorldModel: Exploration-driven Planning with Latent World Models
arXiv:2610.05996v1 Announce Type: cross Abstract: Latent world models based on Joint-Embedding Predictive Architecture (JEPA) are deterministic by design. While successful in fully observable scenarios, this paradigm bre…
- arXiv cs.AIresearchInvestigating Query-Insensitive Behavior in Spatio-Temporal Video Grounding
arXiv:2610.06018v1 Announce Type: cross Abstract: Spatio-temporal video grounding (STVG) aims to localize objects or events described by natural language queries in both space and time. Existing STVG models are typically…