AI Research News Feeds for February 9th, 2026

AI RESEARCH PAPERS & ACADEMIC SOURCES

• EUGens: Efficient, Unified, and General Dense Layers
• Anonymization Prompt Learning for Facial Privacy-Preserving Text-to-Image Generation
• Self-Supervised Video Representation Learning in a Heuristic Decoupled Perspective
• Sketch2Scene: Automatic Generation of Interactive 3D Game Scenes from User's Casual Sketches
• Designing Computational Tools for Exploring Causal Relationships in Qualitative Data
• T$^3$-S2S: Training-free Triplet Tuning for Sketch to Scene Synthesis in Controllable Concept Art Generation
• From Blurry to Believable: Enhancing Low-quality Talking Heads with 3D Generative Priors
• EgoAVU: Egocentric Audio-Visual Understanding
• MGP-KAD: Multimodal Geometric Priors and Kolmogorov-Arnold Decoder for Single-View 3D Reconstruction in Complex Scenes
• Driving with DINO: Vision Foundation Features as a Unified Bridge for Sim-to-Real Generation in Autonomous Driving
• MetaSSP: Enhancing Semi-supervised Implicit 3D Reconstruction through Meta-adaptive EMA and SDF-aware Pseudo-label Evaluation
• M3: High-fidelity Text-to-Image Generation via Multi-Modal, Multi-Agent and Multi-Round Visual Reasoning
• Unsupervised Anomaly Detection of Diseases in the Female Pelvis for Real-Time MR Imaging
• DeDPO: Debiased Direct Preference Optimization for Diffusion Models
• DroneKey++: A Size Prior-free Method and New Benchmark for Drone 3D Pose Estimation from Sequential Images
• ForeHOI: Feed-forward 3D Object Reconstruction from Daily Hand-Object Interaction Videos
• An Interpretable Vision Transformer as a Fingerprint-Based Diagnostic Aid for Kabuki and Wiedemann-Steiner Syndromes
• MMEarth-Bench: Global Model Adaptation via Multimodal Test-Time Training
• Unsupervised MRI-US Multimodal Image Registration with Multilevel Correlation Pyramidal Optimization
• Adaptive and Balanced Re-initialization for Long-timescale Continual Test-time Domain Adaptation
• Halt the Hallucination: Decoupling Signal and Semantic OOD Detection Based on Cascaded Early Rejection
• Taming SAM3 in the Wild: A Concept Bank for Open-Vocabulary Segmentation
• SPDA-SAM: A Self-prompted Depth-Aware Segment Anything Model for Instance Segmentation
• Uncertainty-Aware 4D Gaussian Splatting for Monocular Occluded Human Rendering
• FlowConsist: Make Your Flow Consistent with Real Trajectory
• Robust Pedestrian Detection with Uncertain Modality
• POINTS-GUI-G: GUI-Grounding Journey
• MeDocVL: A Visual Language Model for Medical Document Understanding and Parsing
• A neuromorphic model of the insect visual system for natural image processing
• Point Virtual Transformer
• Learning Human Visual Attention on 3D Surfaces through Geometry-Queried Semantic Priors
• Alleviating Sparse Rewards by Modeling Step-Wise and Long-Term Sampling Effects in Flow-Based GRPO
• POPL-KF: A Pose-Only Geometric Representation-Based Kalman Filter for Point-Line-Based Visual-Inertial Odometry
• Bridging the Indoor-Outdoor Gap: Vision-Centric Instruction-Guided Embodied Navigation for the Last Meters
• ChatUMM: Robust Context Tracking for Conversational Interleaved Generation
• What Is Wrong with Synthetic Data for Scene Text Recognition? A Strong Synthetic Engine with Diverse Simulations and Self-Evolution
• Exploring Specular Reflection Inconsistency for Generalizable Face Forgery Detection
• LAB-Det: Language as a Domain-Invariant Bridge for Training-Free One-Shot Domain Generalization in Object Detection
• Instance-Free Domain Adaptive Object Detection
• Rebenchmarking Unsupervised Monocular 3D Occupancy Prediction
• DreamHome-Pano: Design-Aware and Conflict-Free Panoramic Interior Generation
• FloorplanVLM: A Vision-Language Model for Floorplan Vectorization
• DriveWorld-VLA: Unified Latent-Space World Modeling with Vision-Language-Action for Autonomous Driving
• MicroBi-ConvLSTM: An Ultra-Lightweight Efficient Model for Human Activity Recognition on Resource Constrained Devices
• AdaptOVCD: Training-Free Open-Vocabulary Remote Sensing Change Detection via Adaptive Information Fusion
• Universal Anti-forensics Attack against Image Forgery Detection via Multi-modal Guidance
• An Integer Linear Programming Approach to Geometrically Consistent Partial-Partial Shape Matching
• CauCLIP: Bridging the Sim-to-Real Gap in Surgical Video Understanding via Causality-Inspired Vision-Language Modeling
• PlanViz: Evaluating Planning-Oriented Image Generation and Editing for Computer-Use Tasks
• Can We Build a Monolithic Model for Fake Image Detection? SICA: Semantic-Induced Constrained Adaptation for Unified-Yet-Discriminative Artifact Feature Space Reconstruction
• Clinical-Prior Guided Multi-Modal Learning with Latent Attention Pooling for Gait-Based Scoliosis Screening
• Machine Learning for Detection and Severity Estimation of Sweetpotato Weevil Damage in Field and Lab Conditions
• A Unified Formula for Affine Transformations between Calibrated Cameras
• GaussianPOP: Principled Simplification Framework for Compact 3D Gaussian Splatting via Error Quantification
• Parameters as Experts: Adapting Vision Models with Dynamic Parameter Routing
• RFDM: Residual Flow Diffusion Model for Efficient Causal Video Editing
• Prompt Reinjection: Alleviating Prompt Forgetting in Multimodal Diffusion Transformers
• Seeing Beyond Redundancy: Task Complexity's Role in Vision Token Specialization in VLLMs
• CineScene: Implicit 3D as Effective Scene Representation for Cinematic Video Generation
• MedMO: Grounding and Understanding Multimodal Large Language Model for Medical Images
• COSMOS: Coherent Supergaussian Modeling with Spatial Priors for Sparse-View 3D Splatting
• ALIEN: Analytic Latent Watermarking for Controllable Generation
• Zero-shot Multi-Contrast Brain MRI Registration by Intensity Randomizing T1-weighted MRI (LUMIR25)
• AS-Mamba: Asymmetric Self-Guided Mamba Decoupled Iterative Network for Metal Artifact Reduction
• MultiGraspNet: A Multitask 3D Vision Model for Multi-gripper Robotic Grasping
• Think Proprioceptively: Embodied Visual Reasoning for VLA Manipulation
• Orientation-Robust Latent Motion Trajectory Learning for Annotation-free Cardiac Phase Detection in Fetal Echocardiography
• 3D Object Detection for Autonomous Driving: A Survey
• Nonparametric Evaluation of Noisy ICA Solutions
• STAG: Structural Test-time Alignment of Gradients for Online Adaptation
• Sampling for Model Predictive Trajectory Planning in Autonomous Driving using Normalizing Flows
• Predicting the fatigue life of asphalt concrete using neural networks
• Science-Informed Design of Deep Learning With Applications to Wireless Systems: A Tutorial
• Ensemble Transport Filter via Optimized Maximum Mean Discrepancy
• Relevance-aware Multi-context Contrastive Decoding for Retrieval-augmented Visual Question Answering
• CAST: Character-and-Scene Episodic Memory for Agents
• PersonaPlex: Voice and Role Control for Full Duplex Conversational Speech Models
• What Is Novel? A Knowledge-Driven Framework for Bias-Aware Literature Originality Evaluation
• Quantifying and Attributing Polarization to Annotator Groups
• Uncertainty Drives Social Bias Changes in Quantized Large Language Models
• BenchMarker: An Education-Inspired Toolkit for Highlighting Flaws in Multiple-Choice Benchmarks
• Is my model "mind blurting"? Interpreting the dynamics of reasoning tokens with Recurrence Quantification Analysis (RQA)
• VowelPrompt: Hearing Speech Emotions from Text via Vowel-level Prosodic Augmentation
• RoPE-LIME: RoPE-Space Locality + Sparse-K Sampling for Efficient LLM Attribution
• Judging What We Cannot Solve: A Consequence-Based Approach for Oracle-Free Evaluation of Research-Level Math
• Lost in Speech: Benchmarking, Evaluation, and Parsing of Spoken Code-Switching Beyond Standard UD Assumptions
• Cost-Aware Model Selection for Text Classification: Multi-Objective Trade-offs Between Fine-Tuned Encoders and LLM Prompting in Production
• ReBeCA: Unveiling Interpretable Behavior Hierarchy behind the Iterative Self-Reflection of Language Models with Causal Analysis
• FMBench: Adaptive Large Language Model Output Formatting
• On the Wings of Imagination: Conflicting Script-based Multi-role Framework for Humor Caption Generation
• Evaluating an evidence-guided reinforcement learning framework in aligning light-parameter large language models with decision-making cognition in psychiatric clinical reasoning
• RelayGen: Intra-Generation Model Switching for Efficient Reasoning
• Baichuan-M3: Modeling Clinical Inquiry for Reliable Medical Decision-Making
• Echoes as Anchors: Probabilistic Costs and Attention Refocusing in LLM Reasoning
• Do Prompts Guarantee Safety? Mitigating Toxicity from LLM Generations through Subspace Intervention
• FairJudge: An Adaptive, Debiased, and Consistent LLM-as-a-Judge
• Reading Between the Waves: Robust Topic Segmentation Using Inter-Sentence Audio Features
• Beyond Static Alignment: Hierarchical Policy Control for LLM Safety via Risk-Aware Chain-of-Thought
• Evaluating Prompt Engineering Strategies for Sentiment Control in AI-Generated Texts
• Table-as-Search: Formulate Long-Horizon Agentic Information Seeking as Table Completion
• R-Align: Enhancing Generative Reward Models through Rationale-Centric Meta-Judging
• Visual Word Sense Disambiguation with CLIP through Dual-Channel Text Prompting and Image Augmentations
• SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks
• DAWN: Dependency-Aware Fast Inference for Diffusion LLMs
• STACodec: Semantic Token Assignment for Balancing Acoustic Fidelity and Semantic Information in Audio Codecs
• PhenoLIP: Integrating Phenotype Ontology Knowledge into Medical Vision-Language Pretraining
• Displacement-Resistant Extensions of DPO with Nonconvex $f$-Divergences
• Rare Event Analysis of Large Language Models
• FlowDA: Accurate, Low-Latency Weather Data Assimilation via Flow Matching
• Calibrating Tabular Anomaly Detection via Optimal Transport
• Learning Deep Hybrid Models with Sharpness-Aware Minimization
• Improved Sampling Schedules for Discrete Diffusion Models
• Designing a Robust, Bounded, and Smooth Loss Function for Improved Supervised Learning
• T-STAR: A Context-Aware Transformer Framework for Short-Term Probabilistic Demand Forecasting in Dock-Based Shared Micro-Mobility
• Decoupling Variance and Scale-Invariant Updates in Adaptive Gradient Descent for Unified Vector and Matrix Optimization
• Vision Transformer Finetuning Benefits from Non-Smooth Components
• A Cycle-Consistent Graph Surrogate for Full-Cycle Left Ventricular Myocardial Biomechanics
• Sample Complexity of Causal Identification with Temporal Heterogeneity
• Parameter-free Dynamic Regret: Time-varying Movement Costs, Delayed Feedback, and Memory
• A first realization of reinforcement learning-based closed-loop EEG-TMS
• Revisiting the Generic Transformer: Deconstructing a Strong Baseline for Time Series Foundation Models
• Robustness Beyond Known Groups with Low-rank Adaptation
• Continuous-time reinforcement learning: ellipticity enables model-free value function approximation
• When RL Meets Adaptive Speculative Training: A Unified Training-Serving System
• From Core to Detail: Unsupervised Disentanglement with Entropy-Ordered Flows
• Improving Credit Card Fraud Detection with an Optimized Explainable Boosting Machine
• Deep Unfolded Fractional Optimization for Maximizing Robust Throughput in 6G Networks
• Deep networks learn to parse uniform-depth context-free languages from local statistics
• PackInfer: Compute- and I/O-Efficient Attention for Batched LLM Inference
• Canzona: A Unified, Asynchronous, and Load-Balanced Framework for Distributed Matrix-based Optimizers
• Algebraic Robustness Verification of Neural Networks
• Warm Starts, Cold States: Exploiting Adiabaticity for Variational Ground-States
• Know Your Scientist: KYC as Biosecurity Infrastructure
• Cross-Modal Redundancy and the Geometry of Vision-Language Embeddings
• Inheritance Between Feedforward and Convolutional Networks via Model Projection
• MPIB: A Benchmark for Medical Prompt Injection Attacks and Clinical Safety in LLMs
• Time-uniform conformal and PAC prediction
• High-Dimensional Limit of Stochastic Gradient Flow via Dynamical Mean-Field Theory
• AdFL: In-Browser Federated Learning for Online Advertisement
• Envy-Free Allocation of Indivisible Goods via Noisy Queries
• Advances in Battery Energy Storage Management: Control and Economic Synergies
• A Multiplicative Neural Network Architecture: Locality and Regularity of Appriximation
• HyQuRP: Hybrid quantum-classical neural network with rotational and permutational equivariance for 3D point clouds
• Stopping Computation for Converged Tokens in Masked Diffusion-LM Decoding
• Diffusion-State Policy Optimization for Masked Diffusion Language Models
• Forest canopy height estimation from satellite RGB imagery using large-scale airborne LiDAR-derived training data and monocular depth estimation
• AlertBERT: A noise-robust alert grouping framework for simultaneous cyber attacks
• Operationalizing Stein's Method for Online Linear Optimization: CLT-Based Optimal Tradeoffs
• NECromancer: Breathing Life into Skeletons via BVH Animation
• Evolving Ranking Functions for Canonical Blow-Ups in Positive Characteristic
• Reinforcement Learning-Based Dynamic Management of Structured Parallel Farm Skeletons on Serverless Platforms
• Inference-Time Rethinking with Latent Thought Vectors for Math Reasoning
• Confundo: Learning to Generate Robust Poison for Practical RAG Systems
• Infinite-dimensional generative diffusions via Doob's h-transform
• CytoCrowd: A Multi-Annotator Benchmark Dataset for Cytology Image Analysis
• Makespan Minimization in Split Learning: From Theory to Practice
• Quantum Attention by Overlap Interference: Predicting Sequences from Classical and Many-Body Quantum Data
• Taipan: A Query-free Transfer-based Multiple Sensitive Attribute Inference Attack Solely from Publicly Released Graphs
• Missing At Random as Covariate Shift: Correcting Bias in Iterative Imputation
• Fair Transit Stop Placement: A Clustering Perspective and Beyond
• Revisiting Emotions Representation for Recognition in the Wild
• Optimal Learning-Rate Schedules under Functional Scaling Laws: Power Decay and Warmup-Stable-Decay
• RAIGen: Rare Attribute Identification in Text-to-Image Generative Models
• RanSOM: Second-Order Momentum with Randomized Scaling for Constrained and Unconstrained Optimization
• Are Deep Learning Based Hybrid PDE Solvers Reliable? Why Training Paradigms and Update Strategies Matter
• Uncovering Cross-Objective Interference in Multi-Objective Alignment
• Automatic Detection and Analysis of Singing Mistakes for Music Pedagogy
• Reciprocal Latent Fields for Precomputed Sound Propagation
• Reliable Mislabel Detection for Video Capsule Endoscopy Data
• Optimal Derivative Feedback Control for an Active Magnetic Levitation System: An Experimental Study on Data-Driven Approaches
• A Multi-Token Coordinate Descent Method for Semi-Decentralized Vertical Federated Learning
• Forecasting with Hyper-Trees
• Structural Enforcement of Statistical Rigor in AI-Driven Discovery: A Functional Architecture
• Testing Storage-System Correctness: Challenges, Fuzzing Limitations, and AI-Augmented Opportunities
• Agentic Workflow Using RBA$_\theta$ for Event Prediction
• Toward Faithful and Complete Answer Construction from a Single Document
• Pragmatic Curiosity: A Hybrid Learning-Optimization Paradigm via Active Inference
• Private and interpretable clinical prediction with quantum-inspired tensor train models
• Compressing LLMs with MoP: Mixture of Pruners
• Tempora: Characterising the Time-Contingent Utility of Online Test-Time Adaptation
• Flow Matching for Offline Reinforcement Learning with Discrete Actions
• Optimistic Training and Convergence of Q-Learning -- Extended Version
• MoSE: Mixture of Slimmable Experts for Efficient and Adaptive Language Models
• Latent Structure Emergence in Diffusion Models via Confidence-Based Filtering
• SCONE: A Practical, Constraint-Aware Plug-in for Latent Encoding in Learned DNA Storage
• To 2:4 Sparsity and Beyond: Neuron-level Activation Function to Accelerate LLM Pre-Training
• $f$-FUM: Federated Unlearning via min--max and $f$-divergence
• Provably avoiding over-optimization in Direct Preference Optimization without knowing the data distribution
• A Fast and Generalizable Fourier Neural Operator-Based Surrogate for Melt-Pool Prediction in Laser Processing
• Adaptive Sparse M\"obius Transforms for Learning Polynomials
• On Randomized Algorithms in Online Strategic Classification
• Swap Regret Minimization Through Response-Based Approachability
• PurSAMERE: Reliable Adversarial Purification via Sharpness-Aware Minimization of Expected Reconstruction Error
• Statistical Learning from Attribution Sets
• SOCKET: SOft Collison Kernel EsTimator for Sparse Attention
• How (Not) to Hybridize Neural and Mechanistic Models for Epidemiological Forecasting
• Online Adaptive Reinforcement Learning with Echo State Networks for Non-Stationary Dynamics
• Don't Break the Boundary: Continual Unlearning for OOD Detection Based on Free Energy Repulsion
• Adversarial Learning in Games with Bandit Feedback: Logarithmic Pure-Strategy Maximin Regret
• Enhance and Reuse: A Dual-Mechanism Approach to Boost Deep Forest for Label Distribution Learning
• Evaluating LLM-persona Generated Distributions for Decision-making
• Uniform Spectral Growth and Convergence of Muon in LoRA-Style Matrix Factorization
• Near-Optimal Regret for Distributed Adversarial Bandits: A Black-Box Approach
• EEG Emotion Classification Using an Enhanced Transformer-CNN-BiLSTM Architecture with Dual Attention Mechanisms
• Adaptive Protein Tokenization
• Beyond Code Contributions: How Network Position, Temporal Bursts, and Code Review Activities Shape Contributor Influence in Large-Scale Open Source Ecosystems
• Reclaiming First Principles: A Differentiable Framework for Conceptual Hydrologic Models
• Is Gradient Ascent Really Necessary? Memorize to Forget for Machine Unlearning
• BrokenBind: Universal Modality Exploration beyond Dataset Boundaries
• On the Plasticity and Stability for Post-Training Large Language Models
• The Window Dilemma: Why Concept Drift Detection is Ill-Posed
• Achieving Better Local Regret Bound for Online Non-Convex Bilevel Optimization
• Towards Generalizable Reasoning: Group Causal Counterfactual Policy Optimization for LLM Reasoning
• Adaptive Uncertainty-Aware Tree Search for Robust Reasoning
• Can Microcanonical Langevin Dynamics Leverage Mini-Batch Gradient Noise?
• Evolutionary Generation of Multi-Agent Systems
• Topography scanning as a part of process monitoring in power cable insulation process
• Live Knowledge Tracing: Real-Time Adaptation using Tabular Foundation Models
• Refining the Information Bottleneck via Adversarial Information Separation
• Fine-Grained Model Merging via Modular Expert Recombination
• Learning to Allocate Resources with Censored Feedback
• Degradation of Feature Space in Continual Learning
• DiTS: Multimodal Diffusion Transformers Are Time Series Forecasters
• The hidden risks of temporal resampling in clinical reinforcement learning
• Adaptive-CaRe: Adaptive Causal Regularization for Robust Outcome Prediction
• Pruning at Initialisation through the lens of Graphon Limit: Convergence, Expressivity, and Generalisation
• Memory-Conditioned Flow-Matching for Stable Autoregressive PDE Rollouts
• NanoQuant: Efficient Sub-1-Bit Quantization of Large Language Models
• Diffeomorphism-Equivariant Neural Networks
• Explaining Grokking in Transformers through the Lens of Inductive Bias
• Disentanglement by means of action-induced representations
• Soft Forward-Backward Representations for Zero-shot Reinforcement Learning with General Utilities
• Calibrating Generative AI to Produce Realistic Essays for Data Augmentation
• On the Convergence of Multicalibration Gradient Boosting
• Robust Online Learning
• Weisfeiler and Lehman Go Categorical
• Optimal Abstractions for Verifying Properties of Kolmogorov-Arnold Networks (KANs)
• Gold Exploration using Representations from a Multispectral Autoencoder
• A Unified Framework for LLM Watermarks
• AEGIS: Adversarial Target-Guided Retention-Data-Free Robust Concept Erasure from Diffusion Models
• Next-generation cyberattack detection with large language models: anomaly analysis across heterogeneous logs
• Generating Data-Driven Reasoning Rubrics for Domain-Adaptive Reward Modeling
• On the Identifiability of Steering Vectors in Large Language Models
• SuReNav: Superpixel Graph-based Constraint Relaxation for Navigation in Over-constrained Environments
• Bridging 6G IoT and AI: LLM-Based Efficient Approach for Physical Layer's Optimization Tasks
• AI-Generated Music Detection in Broadcast Monitoring
• AEGPO: Adaptive Entropy-Guided Policy Optimization for Diffusion Models
• The Representational Geometry of Number
• Rethinking Multi-Condition DiTs: Eliminating Redundant Attention via Position-Alignment and Keyword-Scoping
• The Quantum Sieve Tracer: A Hybrid Framework for Layer-Wise Activation Tracing in Large Language Models
• Zero-shot Generalizable Graph Anomaly Detection with Mixture of Riemannian Experts
• TraceCoder: A Trace-Driven Multi-Agent Framework for Automated Debugging of LLM-Generated Code
• NanoFLUX: Distillation-Driven Compression of Large Text-to-Image Generation Models for Mobile Devices
• Supercharging Simulation-Based Inference for Bayesian Optimal Experimental Design
• TamperBench: Systematically Stress-Testing LLM Safety Under Fine-Tuning and Tampering
• PANC: Prior-Aware Normalized Cut for Object Segmentation
• Halluverse-M^3: A multitask multilingual benchmark for hallucination in LLMs
• From Kepler to Newton: Inductive Biases Guide Learned World Models in Transformers
• Implementing Grassroots Logic Programs with Multiagent Transition Systems and AI
• Cochain Perspectives on Temporal-Difference Signals for Learning Beyond Markov Dynamics
• Endogenous Resistance to Activation Steering in Language Models
• Optimal Turkish Subword Strategies at Scale: Systematic Evaluation of Data, Vocabulary, Morphology Interplay
• DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos
• InftyThink+: Effective and Efficient Infinite-Horizon Reasoning via Reinforcement Learning
• Learning a Generative Meta-Model of LLM Activations
• A computational framework for human values
• Conversational Intent-Driven GraphRAG: Enhancing Multi-Turn Dialogue Systems through Adaptive Dual-Retrieval of Flow Patterns and Context Semantics
• Human-AI Co-Embodied Intelligence for Scientific Experimentation and Manufacturing
• Leveraging Spreading Activation for Improved Document Retrieval in Knowledge-Graph-Based RAG Systems
• Yunjue Agent Tech Report: A Fully Reproducible, Zero-Start In-Situ Self-Evolving Agent System for Open-Ended Tasks
• How does information access affect LLM monitors' ability to detect sabotage?
• Bayesian Matrix Decomposition and Applications
• EEG-MACS: Manifold Attention and Confidence Stratification for EEG-based Cross-Center Brain Disease Diagnosis under Unreliable Annotations
• Hyperbolic Fine-Tuning for Large Language Models
• ExpressivityBench: Can LLMs Communicate Implicitly?
• Learning Metal Microstructural Heterogeneity through Spatial Mapping of Diffraction Latent Space Features
• Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks
• SVRepair: Structured Visual Reasoning for Automated Program Repair
• NanoNet: Parameter-Efficient Learning with Label-Scarce Supervision for Lightweight Text Mining Model
• Coding Agents with Environment Interaction: A Theoretical Perspective
• Urban Spatio-Temporal Foundation Models for Climate-Resilient Housing: Scaling Diffusion Transformers for Disaster Risk Prediction
• Self-Improving World Modelling with Latent Actions
• Hear You in Silence: Designing for Active Listening in Human Interaction with Conversational Agents Using Context-Aware Pacing
• Protean Compiler: An Agile Framework to Drive Fine-grain Phase Ordering
• Stop the Flip-Flop: Context-Preserving Verification for Fast Revocable Diffusion Decoding
• Optimal rates for density and mode estimation with expand-and-sparsify representations
• Generics in science communication: Misaligned interpretations across laypeople, scientists, and large language models
• Personagram: Bridging Personas and Product Design for Creative Ideation with Multimodal LLMs
• AnyThermal: Towards Learning Universal Representations for Thermal Perception
• Learning Rate Scaling across LoRA Ranks and Transfer to Full Finetuning
• Multi-Way Representation Alignment
• Emergent Low-Rank Training Dynamics in MLPs with Smooth Activations
• Addressing the Waypoint-Action Gap in End-to-End Autonomous Driving via Vehicle Motion Models
• Coupled Local and Global World Models for Efficient First Order RL
• SR4-Fit: An Interpretable and Informative Classification Algorithm Applied to Prediction of U.S. House of Representatives Elections
• RuleSmith: Multi-Agent LLMs for Automated Game Balancing
• ATEX-CF: Attack-Informed Counterfactual Explanations for Graph Neural Networks
• REBEL: Hidden Knowledge Recovery via Evolutionary-Based Evaluation Loop
• ASMa: Asymmetric Spatio-temporal Masking for Skeleton Action Representation Learning
• Steering Safely or Off a Cliff? Rethinking Specificity and Robustness in Inference-Time Interventions
• GRP-Obliteration: Unaligning LLMs With a Single Unlabeled Prompt
• Can One-sided Arguments Lead to Response Change in Large Language Models?
• Toward generative machine learning for boosting ensembles of climate simulations
• Accelerating Vision Transformers on Brain Processing Unit
• The Condensate Theorem: Transformers are O(n), Not $O(n^2)$
• Can Post-Training Transform LLMs into Causal Reasoners?
• Action Hallucination in Generative Visual-Language-Action Models
• Zero-Trust Runtime Verification for Agentic Payment Protocols: Mitigating Replay and Context-Binding Failures in AP2
• Di3PO -- Diptych Diffusion DPO for Targeted Improvements in Image
• SHINE: A Scalable In-Context Hypernetwork for Mapping Context to LoRA in a Single Pass
• Training Data Selection with Gradient Orthogonality for Efficient Domain Adaptation
• Revisiting Salient Object Detection from an Observer-Centric Perspective
• Generating High-quality Privacy-preserving Synthetic Data
• Empirical Analysis of Adversarial Robustness and Explainability Drift in Cybersecurity Classifiers
• ARIS-RSMA Enhanced ISAC System: Joint Rate Splitting and Beamforming Design
• TFusionOcc: Student's t-Distribution Based Object-Centric Multi-Sensor Fusion Framework for 3D Occupancy Prediction
• Investigating the structure of emotions by analyzing similarity and association of emotion words
• A methodology for analyzing financial needs hierarchy from social discussions using LLM
• TrailBlazer: History-Guided Reinforcement Learning for Black-Box LLM Jailbreaking
• TrajAD: Trajectory Anomaly Detection for Trustworthy LLM Agents
• CORE: Comprehensive Ontological Relation Evaluation for Large Language Models
• Principle-Evolvable Scientific Discovery via Uncertainty Minimization
• Improve Large Language Model Systems with User Logs
• Revisiting the Shape Convention of Transformer Language Models
• Prism: Spectral Parameter Sharing for Multi-Agent Reinforcement Learning
• Efficient-LVSM: Faster, Cheaper, and Better Large View Synthesis Model via Decoupled Co-Refinement Attention
• Completing Missing Annotation: Multi-Agent Debate for Accurate and Scalable Relevant Assessment for IR Benchmarks
• MTQE.en-he: Machine Translation Quality Estimation for English-Hebrew
• Malicious Agent Skills in the Wild: A Large-Scale Security Empirical Study
• Dynamics-Aligned Shared Hypernetworks for Zero-Shot Actuator Inversion
• LIBERO-X: Robustness Litmus for Vision-Language-Action Models
• Which Graph Shift Operator? A Spectral Answer to an Empirical Question
• SPARC: Separating Perception And Reasoning Circuits for Test-time Scaling of VLMs
• Transformer-based Parameter Fitting of Models derived from Bloch-McConnell Equations for CEST MRI Analysis
• Perturbing the Phase: Analyzing Adversarial Robustness of Complex-Valued Neural Networks
• Exploring Sparsity and Smoothness of Arbitrary $\ell_p$ Norms in Adversarial Attacks
• Target noise: A pre-training based neural network initialization for efficient high resolution learning
• ProtoQuant: Quantization of Prototypical Parts For General and Fine-Grained Image Classification
• AgentStepper: Interactive Debugging of Software Development Agents
• Personality as Relational Infrastructure: User Perceptions of Personality-Trait-Infused LLM Messaging
• Sample-Efficient Policy Space Response Oracles with Joint Experience Best Response
• Scaling Speech Tokenizers with Diffusion Autoencoders
• The challenge of generating and evolving real-life like synthetic test data without accessing real-world raw data -- a Systematic Review
• DAVE: Distribution-aware Attribution via ViT Gradient Decomposition
• Trust Regions Sell, But Who's Buying? Overlap Geometry as an Alternative Trust Region for Policy Optimization
• Temperature Scaling Attack Disrupting Model Confidence in Federated Learning
• Humanoid Manipulation Interface: Humanoid Whole-Body Manipulation from Robot-Free Demonstrations
• RAPID: Reconfigurable, Adaptive Platform for Iterative Design
• Multimodal Generative Retrieval Model with Staged Pretraining for Food Delivery on Meituan
• Not All Layers Need Tuning: Selective Layer Restoration Recovers Diversity
• compar:IA: The French Government's LLM arena to collect French-language human prompts and preference data
• SaDiT: Efficient Protein Backbone Design via Latent Structural Tokenization and Diffusion Transformers
• F-GRPO: Don't Let Your Policy Learn the Obvious and Forget the Rare
• GhostCite: A Large-Scale Analysis of Citation Validity in the Age of Large Language Models
• Pairwise is Not Enough: Hypergraph Neural Networks for Multi-Agent Pathfinding
• Jackpot: Optimal Budgeted Rejection Sampling for Extreme Actor-Policy Mismatch Reinforcement Learning
• Large Language Model Reasoning Failures
• Do It for HER: First-Order Temporal Logic Reward Specification in Reinforcement Learning (Extended Version)
• Do LLMs Act Like Rational Agents? Measuring Belief Coherence in Probabilistic Decision Making
• Exposing Weaknesses of Large Reasoning Models through Graph Algorithm Problems
• Trifuse: Enhancing Attention-Based GUI Grounding via Multimodal Fusion
• Difficulty-Estimated Policy Optimization
• Unlocking Noisy Real-World Corpora for Foundation Model Pre-Training via Quality-Aware Tokenization
• Intrinsic Stability Limits of Autoregressive Reasoning: Structural Consequences for Long-Horizon Execution
• AgentCPM-Explore: Realizing Long-Horizon Deep Exploration for Edge-Scale Agents
• JADE: Expert-Grounded Dynamic Evaluation for Open-Ended Professional Tasks
• Progress Constraints for Reinforcement Learning in Behavior Trees
• HyPER: Bridging Exploration and Exploitation for Scalable LLM Reasoning with Hypothesis Path Expansion and Reduction
• LogicSkills: A Structured Benchmark for Formal Reasoning in Large Language Models
• AgentCPM-Report: Interleaving Drafting and Deepening for Open-Ended Deep Research
• SeeUPO: Sequence-Level Agentic-RL with Convergence Guarantees
• Same Answer, Different Representations: Hidden instability in VLMs
• Autoregressive Models for Knowledge Graph Generation
• Semantically Labelled Automata for Multi-Task Reinforcement Learning with LTL Instructions
• Towards Understanding What State Space Models Learn About Code
• Wild Guesses and Mild Guesses in Active Concept Learning
• ScaleEnv: Scaling Environment Synthesis from Scratch for Generalist Interactive Tool-Use Agent Training
• POP: Online Structural Pruning Enables Efficient Inference of Large Foundation Models
• LLM Active Alignment: A Nash Equilibrium Perspective
• An Adaptive Differentially Private Federated Learning Framework with Bi-level Optimization
• From Features to Actions: Explainability in Traditional and Agentic AI Systems
• AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents
• Agentic Uncertainty Reveals Agentic Overconfidence
• Git for Sketches: An Intelligent Tracking System for Capturing Design Evolution
• Recontextualizing Famous Quotes for Brand Slogan Generation
• Rethinking Memory Mechanisms of Foundation Agents in the Second Half
• Analyzing Diffusion and Autoregressive Vision Language Models in Multimodal Embedding Space
• iScheduler: Reinforcement Learning-Driven Continual Optimization for Large-Scale Resource Investment Problems
• HQP: Sensitivity-Aware Hybrid Quantization and Pruning for Ultra-Low-Latency Edge AI Inference
• Allocate Marginal Reviews to Borderline Papers Using LLM Comparative Ranking
• Communication Enhances LLMs' Stability in Strategic Thinking
• Transformer-Based Reinforcement Learning for Autonomous Orbital Collision Avoidance in Partially Observable Environments