AI news archive.
10,661 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchFunctional Subspace, where language models can use vector algebra to solve problems
arXiv:2602.01687v3 Announce Type: replace-cross Abstract: Large language models (LLMs) were invented for natural language tasks such as translation, but they have proved that they can perform highly complex functions acr…
- arXiv cs.AIresearchResidualKV: Residual-Based KV Cache Compression for Efficient Long-Context Inference
arXiv:2602.08005v2 Announce Type: replace-cross Abstract: Efficient long-context inference faces two coupled bottlenecks: KV-cache memory grows linearly with context length, while attention computation grows quadraticall…
- arXiv cs.AIresearchPISCO: Precise Video Instance Insertion with Sparse Control
arXiv:2602.08277v3 Announce Type: replace-cross Abstract: The landscape of AI video generation is undergoing a pivotal shift: moving beyond general generation - which relies on exhaustive prompt-engineering and "cherry-p…
- arXiv cs.AIresearchSemantic Chunking and the Entropy of Natural Language
arXiv:2602.13194v3 Announce Type: replace-cross Abstract: Humans and large language models can predict next letter or word from its prior context much better than random guessing, indicating strong redundancy of language…
- arXiv cs.AIresearchUniST-Pred: A Robust Unified Framework for Spatio-Temporal Traffic Forecasting in Transportation Networks Under Disruptions
arXiv:2602.14049v3 Announce Type: replace-cross Abstract: Spatio-temporal traffic forecasting is a core component of intelligent transportation systems, supporting various downstream tasks such as signal control and netw…
- arXiv cs.AIresearchITO: Multi-View Alignment and Training-Time Fusion for Image-Text Pretraining
arXiv:2603.02767v5 Announce Type: replace-cross Abstract: Image--text contrastive pretraining has become a dominant paradigm for visual representation learning, yet existing methods often yield representations that remai…
- arXiv cs.AIresearchMoEless: Efficient MoE LLM Serving with Serverless Experts
arXiv:2603.06350v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) increasingly adopt Mixture-of-Experts (MoE) architectures to scale efficiently under stringent resource constraints. However, MoE's s…
- arXiv cs.AIresearchDelta-K: Boosting Multi-Instance Generation via Cross-Attention Augmentation
arXiv:2603.10210v2 Announce Type: replace-cross Abstract: While Diffusion Models excel in text-to-image synthesis, they frequently suffer from catastrophic concept omission when generating complex multi-instance scenes.…
- arXiv cs.AIresearchExpertGen: Scalable Sim-to-Real Expert Policy Learning from Imperfect Behavior Priors
arXiv:2603.15956v4 Announce Type: replace-cross Abstract: Learning generalizable and robust behavior cloning policies requires large volumes of high-quality robotics data. While human demonstrations (e.g., through teleop…
- arXiv cs.AIresearchLanguage-Conditioned World Modeling for Visual Navigation
arXiv:2603.26741v2 Announce Type: replace-cross Abstract: Goal-conditioned visual navigation has been a long-standing testbed for embodied AI. We study a natural language-conditioned variant, language-conditioned visual…
- arXiv cs.AIresearchTensorHub: Scalable and Elastic Weight Transfer for LLM RL Training
arXiv:2604.09107v2 Announce Type: replace-cross Abstract: Modern LLM reinforcement learning (RL) workloads require a high-performance weight transfer system to scale training across heterogeneous compute resources. Howev…
- arXiv cs.AIresearchVisionFoundry: Teaching VLMs Visual Perception with Synthetic Images
arXiv:2604.09531v2 Announce Type: replace-cross Abstract: Vision-language models (VLMs) still struggle with visual perception tasks such as spatial understanding and viewpoint recognition, largely because natural image d…
- arXiv cs.AIresearchGrounded World Model: Latent Planning with Language Goals
arXiv:2604.11751v2 Announce Type: replace-cross Abstract: World models such as DINO-WM and LeWM specify the goal with an image, which is difficult to obtain in advance for novel tasks. We present the Grounded World Model…
- arXiv cs.AIresearchFinetuning-Free Diffusion Model with Adaptive Constraint Guidance for Inorganic Crystal Structure Generation
arXiv:2604.13354v4 Announce Type: replace-cross Abstract: Generative diffusion models have emerged as powerful tools for the discovery of inorganic crystal structures, yet steering their sampling process toward user-defi…
- arXiv cs.AIresearchNeuroAI and Beyond: Bridging Between Advances in Neuroscience and Artificial Intelligence
arXiv:2604.18637v3 Announce Type: replace-cross Abstract: Neuroscience and Artificial Intelligence (AI) have made impressive progress in recent years but remain only loosely interconnected. Based on a workshop convened b…
- arXiv cs.AIresearchBeyond LoRA vs. Full Fine-Tuning: Gradient-Guided Optimizer Routing for LLM Adaptation
arXiv:2605.07111v3 Announce Type: replace-cross Abstract: Recent literature on fine-tuning Large Language Models highlights a fundamental debate. While Full Fine-Tuning (FFT) provides greater representational plasticity,…
- arXiv cs.AIresearchVideo Understanding Reward Modeling: A Robust Benchmark and Performant Reward Models
arXiv:2605.07872v2 Announce Type: replace-cross Abstract: Multimodal reward models have advanced substantially in text and image domains, yet progress in video understanding reward modeling remains severely limited by th…
- arXiv cs.AIresearchThe Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space
arXiv:2605.09883v3 Announce Type: replace-cross Abstract: As current Multimodal Large Language Models rapidly saturate canonical visual reasoning benchmarks, a key question emerges: do these strong scores genuinely refle…
- arXiv cs.AIresearchSpeech-based Psychological Crisis Assessment using LLMs
arXiv:2605.10027v2 Announce Type: replace-cross Abstract: Psychological support hotlines provide critical support for individuals experiencing mental health emergencies, yet current assessments largely rely on human oper…
- arXiv cs.AIresearchImage AID via continuous-time reinforcement learning
arXiv:2605.13010v2 Announce Type: replace-cross Abstract: We study image inpainting with generative diffusion models. Existing methods typically either train dedicated task-specific models, or adapt a pretrained diffusio…
- arXiv cs.AIresearchDiagnosing Training Inference Mismatch in LLM Reinforcement Learning via a Zero-Mismatch Reference
arXiv:2605.14220v2 Announce Type: replace-cross Abstract: Modern LLM RL systems separate rollout generation from policy optimization. These two stages are expected to produce token probabilities that match exactly. Howev…
- arXiv cs.AIresearchNeuroAtlas: Benchmarking Foundation Models for Clinical EEG and Brain-Computer Interfaces
arXiv:2605.14698v2 Announce Type: replace-cross Abstract: Foundation models (FMs) promise to extract unified representations that generalize across downstream tasks. They have emerged across fields, including electroence…
- arXiv cs.AIresearchLatent Video Prediction for World Modeling: An Evaluation Uncovering Intriguing Favorable Evidence
arXiv:2605.15618v2 Announce Type: replace-cross Abstract: Self-supervised video models are increasingly framed as world models, yet they are still evaluated almost entirely on clean video and reported as a final task sco…
- arXiv cs.AIresearchWhen Does Self-Supervised Learning Transfer to Time-Series Tasks?
arXiv:2605.19462v2 Announce Type: replace-cross Abstract: Self-supervised learning (SSL) assumes that solving pretext tasks on unlabeled data yields representations that transfer effectively across downstream application…
- arXiv cs.AIresearchTrusted Weights, Treacherous Optimizations? Optimization-Triggered Backdoor Attacks on LLMs
arXiv:2605.20641v2 Announce Type: replace-cross Abstract: Inference optimization aims to minimize the latency and resource consumption of LLM inference while preserving output quality, making large-scale deployment pract…
- arXiv cs.AIresearchRethinking Cross-Layer Information Routing in Diffusion Transformers
arXiv:2605.20708v4 Announce Type: replace-cross Abstract: Diffusion Transformers (DiTs) have become a de facto backbone of modern visual generation, and nearly every major axis of their design -- tokenization, attention,…
- arXiv cs.AIresearchUnderstanding Multimodality in Generative Behavioral Cloning
arXiv:2605.22493v2 Announce Type: replace-cross Abstract: Behavioral cloning becomes challenging when the same observation admits several valid actions. We study how generative behavioral-cloning policies represent such…
- arXiv cs.AIresearchAI Security Research Should Better Incentivize Defense Research
arXiv:2605.23448v2 Announce Type: replace-cross Abstract: This work examines an imbalance in artificial intelligence (AI) security research: the field tends to produce more work on attacking AI systems than on defending…
- arXiv cs.AIresearchWTKO-CNN: Deep Learning Reveals Sequence Motifs Distinguishing Wild-Type and Knockout ATAC-seq Peaks
arXiv:2605.24034v2 Announce Type: replace-cross Abstract: Chromatin regulators can alter transcriptional programs by modifying the accessibility of regulatory DNA elements. Understanding how regulatory sequences differ b…
- arXiv cs.AIresearchConcept Unlearning via Cross-Attention Activation Projection for Diffusion Models
arXiv:2605.25765v2 Announce Type: replace-cross Abstract: Existing closed-form methods for concept unlearning in text-to-image diffusion models typically derive editing directions from fixed text embeddings, which may no…
- arXiv cs.AIresearchVisualNeedle: Benchmarking Active Visual Search in Information-Dense Scenes
arXiv:2605.26380v2 Announce Type: replace-cross Abstract: Frontier multimodal large language models (MLLMs) have been reported to achieve over 90\% accuracy on fine-grained perception benchmarks. However, such scores do…
- arXiv cs.AIresearchRA-MoE: Routing-Aligned Fine-Tuning for Multilingual Adaptation of Mixture-of-Experts Models
arXiv:2605.28306v2 Announce Type: replace-cross Abstract: Mixture-of-Experts (MoE) models enable efficient LLM scaling, yet adapting them to non-English downstream tasks remains challenging. Standard multilingual fine-tu…
- arXiv cs.AIresearchStochastic Gradient Descent with Momentum is Algorithmically Stable
arXiv:2605.28517v2 Announce Type: replace-cross Abstract: Stochastic gradient descent with momentum (SGDM) is one of the most widely used optimization algorithms in machine learning. While optimization properties of SGDM…
- arXiv cs.AIresearchTaxDistill: Improving Metagenomic Taxonomic Annotation via Distilled Genomic Foundation Models
arXiv:2605.28868v2 Announce Type: replace-cross Abstract: Metagenomic taxonomic annotation is essential for interpreting complex microbial communities, yet reliable annotation remains challenging under reference database…
- arXiv cs.AIresearchKBF: Knowledge Boundary as Fingerprint for Language Model and Black-Box API Auditing
arXiv:2605.29524v3 Announce Type: replace-cross Abstract: Relay and reseller APIs mediate access to large language models (LLMs), but users cannot directly verify which model serves them. We introduce \name, a black-box…
- arXiv cs.AIresearchFoundation-Preserving Optimization in Generalized Eigenspace
arXiv:2606.00132v2 Announce Type: replace-cross Abstract: While finetuning effectively adapts foundation models to specialized downstream tasks, it can degrade nontarget capabilities acquired during pretraining. Existing…
- arXiv cs.AIresearchMean--Fluctuation Dynamics at the Edge of Stability
arXiv:2606.05326v2 Announce Type: replace-cross Abstract: We study the dynamics of gradient descent in the Edge of Stability regime, where the learning rate is large enough to induce persistent oscillations in the trajec…
- arXiv cs.AIresearchATM: Why Latent World Models Can Fail to Plan
arXiv:2606.09028v2 Announce Type: replace-cross Abstract: Latent world models can achieve accurate latent prediction yet still differ substantially in downstream planning performance. We argue that a key source of this d…
- arXiv cs.AIresearchLongMoE: Longitudinal Multimodal Learning via Trajectory-Aware Mixture-of-Experts
arXiv:2606.09907v2 Announce Type: replace-cross Abstract: Multimodal clinical learning is increasingly important for integrating diverse patient data, including imaging, text, and personalised health records. However, it…
- arXiv cs.AIresearchEigenism: Ethics for a Human-AI Future
arXiv:2606.12420v2 Announce Type: replace-cross Abstract: Our concepts of survival and self-interest were built for single, continuous biological lives. These ideas break down when applied to artificial intelligence, sin…
- arXiv cs.AIresearchExistence Precedes Value: Joint Modeling of Observational Existence and Evolving States in Time Series Forecasting
arXiv:2606.13571v2 Announce Type: replace-cross Abstract: Real-world time series are often highly incomplete and irregular due to sensor dormancy, transmission delays, and event-driven sampling, making reliable forecasti…
- arXiv cs.AIresearchBeyond the Commitment Boundary: Probing Epiphenomenal Chain-of-Thought in Large Reasoning Models
arXiv:2606.13603v2 Announce Type: replace-cross Abstract: Chain-of-thought (CoT) reasoning is the dominant paradigm for inference-time scaling in language models, yet the causal influence of individual steps on the final…
- arXiv cs.AIresearchA Virtuous AI is an Existential Risk
arXiv:2606.13739v2 Announce Type: replace-cross Abstract: This paper examines trade-offs between AI safety and well-being relative to (i) one of the most promising methods for finetuning super-capable AIs, 'Constitutiona…
- arXiv cs.AIresearchTri-Info: Generalizable, Interpretable Failure Prediction for VLA Models via Information Theory
arXiv:2606.19998v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models are increasingly deployed across diverse tasks, yet they remain black boxes whose physical interactions can cause irreversible…
- arXiv cs.AIresearchFlow Annealing Posterior Sampling for Function-Space Regression and Inverse Problems
arXiv:2606.22346v2 Announce Type: replace-cross Abstract: Principled regression for stochastic processes is a long-standing challenge with deep connections to scientific inverse problems. We introduce Flow Annealing Post…
- arXiv cs.AIresearchFrom Dataset Spectral Geometry to Network Weights: A Geometry-Aware Initialization for Sigmoidal MLPs in Image Classification
arXiv:2606.28444v2 Announce Type: replace-cross Abstract: Classical universal approximation theorems (UAT) establish the expressive power of sigmoidal multilayer perceptrons, but they do not specify how the weights shoul…
- arXiv cs.AIresearchAdaptive Reparametrized Time for Score-Based Diffusion Sampling
arXiv:2607.02137v3 Announce Type: replace-cross Abstract: We study timestep allocation for score-based diffusion sampling, where a learned reverse-time dynamics is discretized on a finite grid. Uniform and hand-crafted s…
- arXiv cs.AIresearchFedSPM: Routing-Enabled Federated Learning under Dual Heterogeneity via Semiparametric Mixture
arXiv:2607.04085v2 Announce Type: replace-cross Abstract: Routing-prediction federated learning has emerged as a new paradigm that reframes inter-client heterogeneity as a resource for system-level intelligence: at infer…
- arXiv cs.AIresearchPrompting Image Generators for Training-free Primitive Shape Abstraction
arXiv:2607.05568v2 Announce Type: replace-cross Abstract: Compact primitive abstractions represent 3D shapes with a few geometric primitives while preserving recognizable components. Learned methods depend on their train…
- arXiv cs.AIresearchThe Hitchhiker's Guide to Monoculture: AI Homogenizes Syntax, Not (Necessarily) Semantics
arXiv:2607.13077v3 Announce Type: replace-cross Abstract: Large language models (LLMs) have been widely reported to homogenize human expression and thought. However, I argue that convergence in language need not imply co…