AI news archive.
10,659 articles filed under Research, newest first, from the outlets listed on the sources page.
- arXiv cs.AIresearchXPress: Parallel Refinement for Diffusion Drafters in Speculative Decoding
arXiv:2608.02438v2 Announce Type: replace Abstract: Block-diffusion drafters like dFlash generate an entire block of draft tokens in a single forward pass, drastically reducing the overhead of multiple-token drafting in…
- arXiv cs.AIresearchBroken Symmetry in LLM Refusal: Answer Release Is More Local Than Refusal Restoration
arXiv:2608.15772v2 Announce Type: replace Abstract: When a language model refuses to answer a prompt, it is unclear whether the correct answer is erased from its internal representations, or merely suppressed at the outp…
- arXiv cs.AIresearchLearning Counterfactual World Models for Embodied Reasoning under Partial Observability
arXiv:2609.05834v2 Announce Type: replace Abstract: World models promise a general route to embodied intelligence: learn predictive dynamics once, then reason, plan, and act with them. Increasingly, the representations b…
- arXiv cs.AIresearchDiscovering and Reasoning of Causality in the Hidden World with Large Language Models
arXiv:2402.03941v4 Announce Type: replace-cross Abstract: Revealing hidden causal variables alongside the underlying causal mechanisms is essential to the development of science. Despite the progress in the past decades,…
- arXiv cs.AIresearchBudgeted Prompt-Set Expansion and Refinement for Language Models via Simulation Optimization
arXiv:2404.08164v3 Announce Type: replace-cross Abstract: A prompt is an instruction or description provided by the user to guide a generative language model toward outputs that meet application requirements. When these…
- arXiv cs.AIresearchEstimating Long-term Heterogeneous Dose-response Curve: Generalization Bound Leveraging Optimal Transport Weights
arXiv:2406.19195v4 Announce Type: replace-cross Abstract: Long-term treatment effect estimation is a significant but challenging problem in many applications. Existing methods rely on ideal assumptions, such as no unobse…
- arXiv cs.AIresearchEF-LLM: A Continual Learning-Based Framework for AI-Assisted Automation, Enhanced Sparse Prediction, and Hallucination Detection
arXiv:2411.00852v3 Announce Type: replace-cross Abstract: Accurate prediction helps achieve supply-demand balance in energy systems,thereby supporting operational decision-making and scheduling.Traditional models,which l…
- arXiv cs.AIresearchDo Deepfake Detectors Work in Reality?
arXiv:2502.10920v2 Announce Type: replace-cross Abstract: Deepfakes, particularly those involving faceswap-based manipulations, have sparked significant societal concern due to their increasing realism and potential for…
- arXiv cs.AIresearchCan Multi-modal (reasoning) LLMs work as deepfake detectors?
arXiv:2503.20084v3 Announce Type: replace-cross Abstract: Deepfake detection remains a critical challenge in the era of advanced generative models, particularly as synthetic media becomes more sophisticated. In this stud…
- arXiv cs.AIresearchRobust and Fine-Grained Detection of AI Generated Texts
arXiv:2504.11952v4 Announce Type: replace-cross Abstract: An ideal detection system for machine generated content is supposed to work well on any generator as many more advanced LLMs come into existence day by day. Exist…
- arXiv cs.AIresearchSimple yet Effective Semi-supervised Knowledge Distillation from Vision-Language Models via Dual-Head Optimization
arXiv:2505.07675v3 Announce Type: replace-cross Abstract: Semi-supervised learning (SSL) has emerged as a practical solution for addressing data scarcity challenges by leveraging unlabeled data. Recently, vision-language…
- arXiv cs.AIresearchBenchmarking Classification Tasks of Chest X-ray AI Across Multinational Public and Private Datasets: a retrospective diagnostic accuracy study
arXiv:2505.16027v2 Announce Type: replace-cross Abstract: Chest X-ray (CXR) AI models are increasingly proposed for triage, screening, and rule-out, where the operating point is fixed to miss few abnormalities. Most eval…
- arXiv cs.AIresearchPCoreSet: Effective Active Learning through Knowledge Distillation from Vision-Language Models
arXiv:2506.00910v3 Announce Type: replace-cross Abstract: Knowledge distillation (KD) is a widely used framework for training compact, task-specific models by transferring the knowledge from teacher models. However, its…
- arXiv cs.AIresearchGeneralized Decoupled Learning for Enhancing Open-Vocabulary Dense Perception
arXiv:2508.11256v2 Announce Type: replace-cross Abstract: Dense visual perception tasks have been constrained by their reliance on predefined categories, limiting their applicability in real-world scenarios where visual…
- arXiv cs.AIresearchIntegrating Large Language Models with Human Data to Design Personalized Decoy Nudges
arXiv:2508.12045v3 Announce Type: replace-cross Abstract: Although nudge strategies can influence behavior, their effectiveness varies across individuals, highlighting the need for personalized nudge design. We hypothesi…
- arXiv cs.AIresearchFantastic Pretraining Optimizers and Where to Find Them
arXiv:2509.02046v3 Announce Type: replace-cross Abstract: AdamW has long been the dominant optimizer in language model pretraining, despite numerous claims that alternative optimizers offer 1.4 to 2x speedup. We posit th…
- arXiv cs.AIresearchFediLoRA: Practical Federated Fine-Tuning of Foundation Models Under Missing-Modality Constraints
arXiv:2509.06984v4 Announce Type: replace-cross Abstract: Federated Learning with LoRA fine-tuning offers an efficient and privacy-aware solution for institutions to collaboratively leverage their large datasets to train…
- arXiv cs.AIresearchDiff-Feat: Diffusion-Based Cross-Modal Feature Extraction for Multi-Label Classification
arXiv:2509.15553v2 Announce Type: replace-cross Abstract: Multi-label classification has broad applications and depends on powerful representations capable of capturing multiple interactions. We introduce Diff-Feat, a si…
- arXiv cs.AIresearchHierarchical Alignment: Block-Level LoRA-DPO for Preference Tuning in Large Language Models
arXiv:2510.12044v2 Announce Type: replace-cross Abstract: Preference tuning typically distributes adapter updates across every Transformer layer. We study how the placement of a fixed LoRA budget affects Direct Preferenc…
- arXiv cs.AIresearchThink Parallax: Solving Multi-Hop Problems via Multi-View Knowledge-Graph-Based Retrieval-Augmented Generation
arXiv:2510.15552v5 Announce Type: replace-cross Abstract: Large language models (LLMs) still struggle with multi-hop reasoning over knowledge-graphs (KGs), and we identify a previously overlooked structural reason for th…
- arXiv cs.AIresearchFoundation Models in Medical Image Analysis: A Systematic Review and Meta-Analysis
arXiv:2510.16973v2 Announce Type: replace-cross Abstract: Recent advancements in foundation models (FMs) have catalyzed a paradigm shift in medical image analysis. Unlike traditional task-specific artificial intelligence…
- arXiv cs.AIresearchBeyond Reasoning Gains: Mitigating General-Capability Forgetting in Large Reasoning Models
arXiv:2510.21978v3 Announce Type: replace-cross Abstract: Reinforcement learning with verifiable rewards (RLVR) has delivered impressive gains in mathematical and multimodal reasoning and has become a standard post-train…
- arXiv cs.AIresearchBridging the Divide: End-to-End Sequence-Graph Learning
arXiv:2510.25126v3 Announce Type: replace-cross Abstract: Many real-world prediction tasks, particularly those involving entities such as customers or patients, involve both {sequential} and {relational} data. Each entit…
- arXiv cs.AIresearchEfficient Linear Attention for Multivariate Time Series Modeling via Entropy Calibration
arXiv:2511.03190v2 Announce Type: replace-cross Abstract: Attention mechanisms are effective for multivariate time series modeling because they adaptively aggregate information across timestamps and variables. Their quad…
- arXiv cs.AIresearchCRAFT-E: An Interpretable and Editable Neuro-Symbolic Framework for Affordance-Grounded Active Object Retrieval
arXiv:2512.04231v2 Announce Type: replace-cross Abstract: Assistive robots operating in unstructured environments must reason not only about what objects are, but how they can be used, including non-canonical and improvi…
- arXiv cs.AIresearchA Causal Perspective on the Role of Training Dynamics for Interpreting Privacy Risks in Code Models
arXiv:2512.07814v3 Announce Type: replace-cross Abstract: Large language models for code (LLM4Code) have greatly improved developer productivity but also raise privacy concerns due to their reliance on open-source reposi…
- arXiv cs.AIresearchLLM-Generated Counterfactual Stress Scenarios for Portfolio Risk Simulation via Hybrid Prompt-RAG Pipeline
arXiv:2512.07867v2 Announce Type: replace-cross Abstract: We develop a transparent, auditable large language model (LLM) pipeline for macro-financial stress testing. Structured prompting, with optional retrieval of count…
- arXiv cs.AIresearchOne Leak Away: How Pretrained Model Exposure Amplifies Jailbreak Risks in Finetuned LLMs
arXiv:2512.14751v4 Announce Type: replace-cross Abstract: Finetuning pretrained large language models (LLMs) has become the standard paradigm for developing downstream applications. However, its security implications rem…
- arXiv cs.AIresearchDo Latent Tokens Think? A Causal and Adversarial Analysis of Chain-of-Continuous-Thought
arXiv:2512.21711v2 Announce Type: replace-cross Abstract: Latent tokens are gaining attention for enhancing reasoning in large language models (LLMs), yet their internal mechanisms remain unclear. This paper examines the…
- arXiv cs.AIresearchNavigating Ideation Space: Decomposed Conceptual Representations for Positioning Scientific Ideas
arXiv:2601.08901v2 Announce Type: replace-cross Abstract: Scientific discovery requires new ideas to be situated within an ever-expanding landscape of existing knowledge. An emerging challenge is how to identify conceptu…
- arXiv cs.AIresearchNoiseFormer -- Noise Diffused Symmetric Attention Transformer
arXiv:2601.11619v2 Announce Type: replace-cross Abstract: Transformer architecture has been very successful long runner in the field of Deep Learning (DL) and Large Language Models (LLM) because of its powerful attention…
- arXiv cs.AIresearchGraphDancer: Training LLMs to Explore and Reason over Graphs via Curriculum Reinforcement Learning
arXiv:2602.02518v3 Announce Type: replace-cross Abstract: Large language models (LLMs) increasingly rely on external knowledge to improve factuality, yet many knowledge sources on the Web, such as scholarly indexes and p…
- arXiv cs.AIresearchHow well are open sourced AI-generated image detection models out-of-the-box: A comprehensive benchmark study
arXiv:2602.07814v2 Announce Type: replace-cross Abstract: As AI-generated images proliferate across digital platforms, reliable detection methods have become critical for combating misinformation and maintaining content…
- arXiv cs.AIresearchFlowHOI: Flow-based Semantics-Grounded Generation of Hand-Object Interactions for Dexterous Robot Manipulation
arXiv:2602.13444v2 Announce Type: replace-cross Abstract: Recent vision-language-action (VLA) models can generate plausible end-effector motions, yet they often fail in long-horizon, contact-rich tasks because the underl…
- arXiv cs.AIresearchFlowette: Flow Matching with Graphette Priors for Graph Generation
arXiv:2602.23566v5 Announce Type: replace-cross Abstract: We study generative modeling of graphs with recurring subgraph motifs. We propose Flowette, a continuous flow matching framework that employs a graph neural netwo…
- arXiv cs.AIresearchDirectional Attention Guidance for Diffusion Models via Sparse Hopfield Retrieval
arXiv:2603.02531v3 Announce Type: replace-cross Abstract: Classifier-Free Guidance (CFG) improves sample quality in diffusion models, but its dual-pass inference and null-condition training limit its use in few-step regi…
- arXiv cs.AIresearchActivity Recognition from Smart Insole Sensor Data Using a Circular Dilated CNN
arXiv:2603.04477v2 Announce Type: replace-cross Abstract: Smart insoles equipped with pressure sensors, accelerometers, and gyroscopes offer a non-intrusive way of monitoring human gait and posture. We present an activit…
- arXiv cs.AIresearchDo Compact SSL Backbones Matter for Audio Deepfake Detection? A Controlled Study with RAPTOR
arXiv:2603.06164v2 Announce Type: replace-cross Abstract: Self-supervised learning (SSL) underpins modern audio deepfake detection, yet most prior work centers on a single large wav2vec2-XLSR backbone, leaving compact un…
- arXiv cs.AIresearchCAMEL-CLIP: Channel-aware Multimodal Electroencephalography-text Alignment for Generalizable Brain Foundation Models
arXiv:2603.13272v2 Announce Type: replace-cross Abstract: Electroencephalography (EEG) foundation models have shown promise for learning generalizable representations, yet they remain sensitive to channel heterogeneity,…
- arXiv cs.AIresearchArrayTac: A Closed-Loop Piezoelectric Tactile Platform for Continuously Tunable Rendering of Shape, Stiffness, and Friction
arXiv:2603.13829v3 Announce Type: replace-cross Abstract: Human touch depends on the integration of shape, stiffness, and friction, yet existing tactile displays cannot render these cues together as continuously tunable,…
- arXiv cs.AIresearchHiMu: Hierarchical Multimodal Frame Selection for Long Video Question Answering
arXiv:2603.18558v3 Announce Type: replace-cross Abstract: Long-form video question answering requires reasoning over extended temporal contexts, making frame selection a critical bottleneck for multi-modal large language…
- arXiv cs.AIresearchAutomatic Configuration of LLM Post-Training Pipelines
arXiv:2603.18773v2 Announce Type: replace-cross Abstract: LLM post-training pipelines that combine supervised fine-tuning and reinforcement learning are difficult to configure under realistic compute budgets: the configu…
- arXiv cs.AIresearchGood Scores, Bad Data: A Metric for Multimodal Coherence
arXiv:2603.25924v2 Announce Type: replace-cross Abstract: Multimodal AI systems are evaluated by downstream task accuracy, but high accuracy does not mean the underlying data is coherent. A model can score well on Visual…
- arXiv cs.AIresearchLinearARD: Linear-Memory Attention Distillation for RoPE Restoration
arXiv:2604.00004v3 Announce Type: replace-cross Abstract: The extension of context windows in Large Language Models is typically facilitated by scaling positional encodings followed by Continued Pre-Training (CPT). While…
- arXiv cs.AIresearchTowards Real-world Human Behavior Simulation: Benchmarking Large Language Models on Long-horizon, Cross-scenario, Heterogeneous Behavior Traces
arXiv:2604.08362v3 Announce Type: replace-cross Abstract: The emergence of Large Language Models (LLMs) has illuminated the potential for a general-purpose user simulator. However, existing benchmarks remain constrained…
- arXiv cs.AIresearchAffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic Manipulation
arXiv:2604.11674v4 Announce Type: replace-cross Abstract: Many everyday robot manipulation skills are affordance-dependent, with success determined by whether the robot contacts the functional object region required by t…
- arXiv cs.AIresearchThe Long Delay to Arithmetic Generalization
arXiv:2604.13082v3 Announce Type: replace-cross Abstract: Large language models today still make mistakes on multi-digit arithmetic when they do not have tools, and part of the reason is how numbers are split into tokens…
- arXiv cs.AIresearchKV Packet: Recomputation-Free Context-Independent KV Caching for LLMs
arXiv:2604.13226v3 Announce Type: replace-cross Abstract: Large Language Models (LLMs) rely heavily on Key-Value (KV) caching to minimize inference latency. However, standard KV caches are context-dependent: reusing a ca…
- arXiv cs.AIresearchBeyond Feature Fusion: Contextual Bayesian PEFT for Multimodal Uncertainty Estimation
arXiv:2604.16615v2 Announce Type: replace-cross Abstract: We introduce CoCo-LoRA, a multimodal, uncertainty-aware parameter-efficient fine-tuning method for text prediction tasks accompanied by audio context. Existing PE…
- arXiv cs.AIresearchCross-Attentive Bayesian Low-Rank Adaptation for Multimodal Uncertainty Estimation
arXiv:2604.16657v2 Announce Type: replace-cross Abstract: Parameter-efficient fine-tuning (PEFT) enables efficient adaptation of frozen language models, but most PEFT methods remain deterministic or unimodal, limiting th…