AI Research News Feeds for February 13th, 2026

AI RESEARCH PAPERS & ACADEMIC SOURCES

• Stress Tests REVEAL Fragile Temporal and Visual Grounding in Video-Language Models
• Advancing Digital Twin Generation Through a Novel Simulation Framework and Quantitative Benchmarking
• Selective Prior Synchronization via SYNC Loss
• MDE-VIO: Enhancing Visual-Inertial Odometry Using Learned Depth Priors
• Exploring Real-Time Super-Resolution: Benchmarking and Fine-Tuning for Streaming Content
• ArtContext: Contextualizing Artworks with Open-Access Art History Articles and Wikidata Knowledge through a LoRA-Tuned CLIP Model
• Ctrl&Shift: High-Quality Geometry-Aware Object Manipulation in Visual Generation
• A Dual-Branch Framework for Semantic Change Detection with Boundary and Temporal Awareness
• Arbitrary Ratio Feature Compression via Next Token Prediction
• What if Agents Could Imagine? Reinforcing Open-Vocabulary HOI Comprehension through Generation
• Vascular anatomy-aware self-supervised pre-training for X-ray angiogram analysis
• Supervise-assisted Multi-modality Fusion Diffusion Model for PET Restoration
• LUVE : Latent-Cascaded Ultra-High-Resolution Video Generation with Dual Frequency Experts
• Move What Matters: Parameter-Efficient Domain Adaptation via Optimal Transport Flow for Collaborative Perception
• A Large Language Model for Disaster Structural Reconnaissance Summarization
• Electrostatics-Inspired Surface Reconstruction (EISR): Recovering 3D Shapes as a Superposition of Poisson's PDE Solutions
• GR-Diffusion: 3D Gaussian Representation Meets Diffusion in Whole-Body PET Reconstruction
• EmoSpace: Fine-Grained Emotion Prototype Learning for Immersive Affective Content Generation
• Clutt3R-Seg: Sparse-view 3D Instance Segmentation for Language-grounded Grasping in Cluttered Scenes
• Egocentric Gaze Estimation via Neck-Mounted Camera
• U-Net with Hadamard Transform and DCT Latent Spaces for Next-day Wildfire Spread Prediction
• RI-Mamba: Rotation-Invariant Mamba for Robust Text-to-Shape Retrieval
• TG-Field: Geometry-Aware Radiative Gaussian Fields for Tomographic Reconstruction
• GSO-SLAM: Bidirectionally Coupled Gaussian Splatting and Direct Visual Odometry
• STVG-R1: Incentivizing Instance-Level Reasoning and Grounding in Videos via Reinforcement Learning
• Adaptive Debiasing Tsallis Entropy for Test-Time Adaptation
• Code2Worlds: Empowering Coding LLMs for 4D World Generation
• Light4D: Training-Free Extreme Viewpoint 4D Video Relighting
• Efficient Segment Anything with Depth-Aware Fusion and Limited Training Data
• JEPA-VLA: Video Predictive Embedding is Needed for VLA Models
• WorldTree: Towards 4D Dynamic Worlds from Monocular Video using Tree-Chains
• DiffPlace: Street View Generation via Place-Controllable Diffusion Model Enhancing Place Recognition
• Spatial Chain-of-Thought: Bridging Understanding and Generation Models for Spatial Reasoning Generation
• Can Local Vision-Language Models improve Activity Recognition over Vision Transformers? -- Case Study on Newborn Resuscitation
• Projected Representation Conditioning for High-fidelity Novel View Synthesis
• A DMD-Based Adaptive Modulation Method for High Dynamic Range Imaging in High-Glare Environments
• GigaBrain-0.5M*: a VLA That Learns From World Model-Based Reinforcement Learning
• AssetFormer: Modular 3D Assets Generation with Autoregressive Transformer
• PosterOmni: Generalized Artistic Poster Creation via Task Distillation and Unified Reward Feedback
• FAIL: Flow Matching Adversarial Imitation Learning for Image Generation
• TexSpot: 3D Texture Enhancement with Spatially-uniform Point Latent Representation
• DreamID-Omni: Unified Framework for Controllable Human-Centric Audio-Video Generation
• EO-VAE: Towards A Multi-sensor Tokenizer for Earth Observation Data
• Best of Both Worlds: Multimodal Reasoning and Generation via Unified Discrete Flow Matching
• Stroke of Surprise: Progressive Semantic Illusions in Vector Sketching
• Mitigating Error Accumulation in Continuous Navigation via Memory-Augmented Kalman Filtering
• U-DAVI: Uncertainty-Aware Diffusion-Prior-Based Amortized Variational Inference for Image Reconstruction
• Learning Perceptual Representations for Gaming NR-VQA with Multi-Task FR Signals
• UPDA: Unsupervised Progressive Domain Adaptation for No-Reference Point Cloud Quality Assessment
• Learning A Physical-aware Diffusion Model Based on Transformer for Underwater Image Enhancement
• Scale Contrastive Learning with Selective Attentions for Blind Image Quality Assessment
• Dual Frequency Branch Framework with Reconstructed Sliding Windows Attention for AI-Generated Image Detection
• From Implicit Ambiguity to Explicit Solidity: Diagnosing Interior Geometric Degradation in Neural Radiance Fields for Dense 3D Scene Understanding
• Response-Based Knowledge Distillation for Multilingual Jailbreak Prevention Unwittingly Compromises Safety
• Retrieval Heads are Dynamic
• PRIME: Policy-Reinforced Iterative Multi-agent Execution for Algorithmic Reasoning in Large Language Models
• Synthesizing the Virtual Advocate: A Multi-Persona Speech Generation Framework for Diverse Linguistic Jurisdictions in Indic Languages
• Author-in-the-Loop Response Generation and Evaluation: Integrating Author Expertise and Intent in Responses to Peer Review
• Barriers to Discrete Reasoning with Transformers: A Survey Across Depth, Exactness, and Bandwidth
• Mechanistic Interpretability for Large Language Model Alignment: Progress, Challenges, and Future Directions
• Code Mixologist : A Practitioner's Guide to Building Code-Mixed LLMs
• MetaMem: Evolving Meta-Memory for Knowledge Utilization through Self-Reflective Symbolic Optimization
• Mechanistic Evidence for Faithfulness Decay in Chain-of-Thought Reasoning
• The Automatic Verification of Image-Text Claims (AVerImaTeC) Shared Task
• SurveyLens: A Research Discipline-Aware Benchmark for Automatic Survey Generation
• Are Aligned Large Language Models Still Misaligned?
• Evaluating Alignment of Behavioral Dispositions in LLMs
• Advancing AI Trustworthiness Through Patient Simulation: Risk Assessment of Conversational Agents for Antidepressant Selection
• LoopFormer: Elastic-Depth Looped Transformers for Latent Reasoning via Shortcut Modulation
• ADRD-Bench: A Preliminary LLM Benchmark for Alzheimer's Disease and Related Dementias
• When Audio-LLMs Don't Listen: A Cross-Linguistic Study of Modality Arbitration
• Pretraining A Large Language Model using Distributed GPUs: A Memory-Efficient Decentralized Paradigm
• SIGHT: Reinforcement Learning with Self-Evidence and Information-Gain Diverse Branching for Search Agent
• PRIME: A Process-Outcome Alignment Benchmark for Verifiable Reasoning in Mathematics and Engineering
• Scene-Aware Memory Discrimination: Deciding Which Personal Knowledge Stays
• PACE: Prefix-Protected and Difficulty-Aware Compression for Efficient Reasoning
• Which Feedback Works for Whom? Differential Effects of LLM-Generated Feedback Elements Across Learner Profiles
• Finding Sense in Nonsense with Generated Contexts: Perspectives from Humans and Language Models
• Thinking with Drafting: Optical Decompression via Logical Reconstruction
• Think Longer to Explore Deeper: Learn to Explore In-Context via Length-Incentivized Reinforcement Learning
• A Subword Embedding Approach for Variation Detection in Luxembourgish User Comments
• LLM-based Triplet Extraction from Financial Reports
• Benchmark Illusion: Disagreement among LLMs and Its Scientific Consequences
• Cross-Modal Robustness Transfer (CMRT): Training Robust Speech Translation Models Using Adversarial Text
• Who is the richest club in the championship? Detecting and Rewriting Underspecified Questions Improve QA Performance
• Do Large Language Models Adapt to Language Variation across Socioeconomic Status?
• Scaling Model and Data for Multilingual Machine Translation with Open Large Language Models
• DHPLT: large-scale multilingual diachronic corpora and word representations for semantic change modelling
• Automatic Simplification of Common Vulnerabilities and Exposures Descriptions
• LaCy: What Small Language Models Can and Should Learn is Not Just a Question of Loss
• Disentangling Ambiguity from Instability in Large Language Models: A Clinical Text-to-SQL Case Study
• Composition-RL: Compose Your Verifiable Prompts for Reinforcement Learning of Large Language Models
• P-GenRM: Personalized Generative Reward Model with Test-time User-based Scaling
• A Rule-based Computational Model for Gaidhlig Morphology
• WavBench: Benchmarking Reasoning, Colloquialism, and Paralinguistics for End-to-End Spoken Dialogue Models
• CitiLink-Minutes: A Multilayer Annotated Dataset of Municipal Meeting Minutes
• Query-focused and Memory-aware Reranker for Long Context Processing
• ExStrucTiny: A Benchmark for Schema-Variable Structured Information Extraction from Document Images
• Detecting Overflow in Compressed Token Representations for Retrieval-Augmented Generation
• On-Policy Context Distillation for Language Models
• Althea: Human-AI Collaboration for Fact-Checking and Critical Reasoning
• Agent-Diff: Benchmarking LLM Agents on Enterprise API Tasks via Code Execution with State-Diff-Based Evaluation
• ABot-M0: VLA Foundation Model for Robotic Manipulation with Action Manifold Learning
• Jailbreaking Leaves a Trace: Understanding and Detecting Jailbreak Attacks from Internal Representations of Large Language Models
• Mask What Matters: Mitigating Object Hallucinations in Multimodal Large Language Models with Object-Aligned Visual Contrastive Decoding
• More Haste, Less Speed: Weaker Single-Layer Watermark Improves Distortion-Free Watermark Ensembles
• Artificial intelligence is creating a new global linguistic hierarchy
• Embodied Agents Meet Personalization: Investigating Challenges and Solutions Through the Lens of Memory Utilization
• DD-MDN: Human Trajectory Forecasting with Diffusion-Based Dual Mixture Density Networks and Uncertainty Self-Calibration
• ReTracing: An Archaeological Approach Through Body, Machine, and Generative Systems
• Aggregate Models, Not Explanations: Improving Feature Importance Estimation
• Decentralized Non-convex Stochastic Optimization with Heterogeneous Variance
• How to Sample High Quality 3D Fractals for Action Recognition Pre-Training?
• A Comparative Study of MAP and LMMSE Estimators for Blind Inverse Problems
• EqDeepRx: Learning a Scalable MIMO Receiver
• Free Lunch for Stabilizing Rectified Flow Inversion
• Scale-Invariant Fast Convergence in Games
• DMAP: A Distribution Map for Text
• Echo: Towards Advanced Audio Comprehension via Audio-Interleaved Reasoning
• TADA! Tuning Audio Diffusion Models through Activation Steering
• Insights on Muon from Simple Quadratics
• Benchmarking Vision-Language Models for French PDF-to-Markdown Conversion
• Calibrated Bayesian Deep Learning for Explainable Decision Support Systems Based on Medical Imaging
• The Implicit Bias of Logit Regularization
• Safety Beyond the Training Data: Robust Out-of-Distribution MPC via Conformalized System Level Synthesis
• Iskra: A System for Inverse Geometry Processing
• Towards Personalized Bangla Book Recommendation: A Large-Scale Multi-Entity Book Graph Dataset
• Convex Markov Games and Beyond: New Proof of Existence, Characterization and Learning Algorithms for Nash Equilibria
• Moonshine v2: Ergodic Streaming Encoder ASR for Latency-Critical Speech Applications
• Is Online Linear Optimization Sufficient for Strategic Robustness?
• T3D: Few-Step Diffusion Language Models via Trajectory Self-Distillation with Direct Discriminative Optimization
• MonarchRT: Efficient Attention for Real-Time Video Generation
• Learning to Control: The iUzawa-Net for Nonsmooth Optimal Control of Linear PDEs
• Accelerating nuclear-norm regularized low-rank matrix optimization through Burer-Monteiro decomposition
• Optimizing Sampling Patterns for Compressed Sensing MRI with Diffusion Generative Models
• SeqRisk: Transformer-augmented latent variable model for robust survival prediction with longitudinal data
• Fine-tuning Quantized Neural Networks with Zeroth-order Optimization
• Hyperparameter Transfer with Mixture-of-Expert Layers
• Accelerating Large Language Model Inference with Self-Supervised Early Exits
• Feature-Based Interpretable Surrogates for Optimization
• Controlling Dynamical Systems into Unseen Target States Using Machine Learning
• SonicSieve: Bringing Directional Speech Extraction to Smartphones Using Acoustic Microstructures
• Empirical Likelihood-Based Fairness Auditing: Distribution-Free Certification and Flagging
• Potential-energy gating for robust state estimation in bistable stochastic systems
• DICE: Diffusion Large Language Models Excel at Generating CUDA Kernels
• Dopamine: Brain Modes, Not Brains
• U-Former ODE: Fast Probabilistic Forecasting of Irregular Time Series
• TUBO: A Tailored ML Framework for Reliable Network Traffic Forecasting
• MUSE: Multi-Tenant Model Serving With Seamless Model Updates
• Temperature as a Meta-Policy: Adaptive Temperature in LLM Reinforcement Learning
• Latent-Variable Learning of SPDEs via Wiener Chaos
• Temporal Difference Learning with Constrained Initial Representations
• SpaTeoGL: Spatiotemporal Graph Learning for Interpretable Seizure Onset Zone Analysis from Intracranial EEG
• TopoFair: Linking Topological Bias to Fairness in Link Prediction Benchmarks
• From Path Signatures to Sequential Modeling: Incremental Signature Contributions for Offline RL
• Deep Kernel Fusion for Transformers
• CAAL: Confidence-Aware Active Learning for Heteroscedastic Atmospheric Regression
• Towards Sustainable Investment Policies Informed by Opponent Shaping
• Robust Optimization Approach and Learning Based Hide-and-Seek Game for Resilient Network Design
• A$^{2}$V-SLP: Alignment-Aware Variational Modeling for Disentangled Sign Language Production
• In-Context Function Learning in Large Language Models
• Universal Diffusion-Based Probabilistic Downscaling
• Learning Conditional Averages
• Extending Puzzle for Mixture-of-Experts Reasoning Models with Application to GPT-OSS Acceleration
• Temporally Unified Adversarial Perturbations for Time Series Forecasting
• Using predictive multiplicity to measure individual performance within the AI Act
• Are Two LLMs Better Than One? A Student-Teacher Dual-Head LLMs Architecture for Pharmaceutical Content Optimization
• RAM-Net: Expressive Linear Attention with Selectively Addressable Memory
• Momentum LMS Theory beyond Stationarity: Stability, Tracking, and Regret
• FedGRPO: Privately Optimizing Foundation Models with Group-Relative Rewards from Domain Client
• Improved state mixing in higher-order and block diagonal linear recurrent networks
• Protein Circuit Tracing via Cross-layer Transcoders
• PrefillShare: A Shared Prefill Module for KV Reuse in Multi-LLM Disaggregated Serving
• Improving HPC Code Generation Capability of LLMs via Online Reinforcement Learning with Real-Machine Benchmark Rewards
• PathCRF: Ball-Free Soccer Event Detection via Possession Path Inference from Player Trajectories
• Empirical Gaussian Processes
• Geometry of Uncertainty: Learning Metric Spaces for Multimodal State Estimation in RL
• Few-Shot Design Optimization by Exploiting Auxiliary Information
• Capability-Oriented Training Induced Alignment Risk
• Oscillators Are All You Need: Irregular Time Series Modelling via Damped Harmonic Oscillators with Closed-Form Solutions
• It's TIME: Towards the Next Generation of Time Series Forecasting Benchmarks
• SafeNeuron: Neuron-Level Safety Alignment for Large Language Models
• Amortized Molecular Optimization via Group Relative Policy Optimization
• How Sampling Shapes LLM Alignment: From One-Shot Optima to Iterative Dynamics
• WaveFormer: Wavelet Embedding Transformer for Biomedical Signals
• Learning to Forget Attention: Memory Consolidation for Adaptive Compute Reduction
• Diffusion Alignment Beyond KL: Variance Minimisation as Effective Policy Optimiser
• Categorical Flow Maps
• Community Concealment from Unsupervised Graph Learning-Based Clustering
• Self-Supervised Learning via Flow-Guided Neural Operator on Time-Series Data
• Function-Space Decoupled Diffusion for Forward and Inverse Modeling in Carbon Capture and Storage
• When and What to Ask: AskBench and Rubric-Guided RLVR for LLM Clarification
• Generative AI-Driven Phase Control for RIS-Aided Cell-Free Massive MIMO Systems
• Active Zero: Self-Evolving Vision-Language Models through Active Environment Exploration
• Unlearnable phases of matter
• Hierarchical Testing of a Hybrid Machine Learning-Physics Global Atmosphere Model
• Amortised and provably-robust simulation-based inference
• Sample-Free Safety Assessment of Neural Network Controllers via Taylor Methods
• Traffic Flow Reconstruction from Limited Collected Data
• Latent Forcing: Reordering the Diffusion Trajectory for Pixel-Space Image Generation
• The Cost of Learning under Multiple Change Points
• Optimizing Agent Planning for Security and Autonomy
• Surface impedance inference via neural fields and sparse acoustic data obtained by a compact array
• Adaptive Power Iteration Method for Differentially Private PCA
• Calibration and Evaluation of Car-Following Models for Autonomous Shuttles Using a Novel Multi-Criteria Framework
• HyperDet: 3D Object Detection with Hyper 4D Radar Point Clouds
• PLESS: Pseudo-Label Enhancement with Spreading Scribbles for Weakly Supervised Segmentation
• Enforcing Reciprocity in Operator Learning for Seismic Wave Propagation
• LAER-MoE: Load-Adaptive Expert Re-layout for Efficient Mixture-of-Experts Training
• Estimation of instrument and noise parameters for inverse problem based on prior diffusion model
• PAC-Bayesian Generalization Guarantees for Fairness on Stochastic and Deterministic Classifiers
• NarraScore: Bridging Visual Narrative and Musical Dynamics via Hierarchical Affective Control
• STaR: Scalable Task-Conditioned Retrieval for Long-Horizon Multimodal Robot Memory
• GAC-KAN: An Ultra-Lightweight GNSS Interference Classifier for GenAI-Powered Consumer Edge Devices
• Predicting the post-wildfire mudflow onset using machine learning models on multi-parameter experimental data
• AM-FM: A Foundation Model for Ambient Intelligence Through WiFi
• Adaptive Physics Transformer with Fused Global-Local Attention for Subsurface Energy Systems
• Towards Compressive and Scalable Recurrent Memory
• Charting Empirical Laws for LLM Fine-Tuning in Scientific Multi-Discipline Learning
• Protein Language Model Embeddings Improve Generalization of Implicit Transfer Operators
• The Magic Correlations: Understanding Knowledge Transfer from Pretraining to Supervised Fine-Tuning
• Patch the Distribution Mismatch: RL Rewriting Agent for Stable Off-Policy SFT
• Learning Glioblastoma Tumor Heterogeneity Using Brain Inspired Topological Neural Networks
• Evaluating Memory Structure in LLM Agents
• Efficient Analysis of the Distilled Neural Tangent Kernel
• Structured Hybrid Mechanistic Models for Robust Estimation of Time-Dependent Intervention Outcomes
• Toward Adaptive Non-Intrusive Reduced-Order Models: Design and Challenges
• WSBD: Freezing-Based Optimizer for Quantum Neural Networks
• Provably Efficient Algorithms for S- and Non-Rectangular Robust MDPs with General Parameterization
• Sparse Semantic Dimension as a Generalization Certificate for LLMs
• CADET: Context-Conditioned Ads CTR Prediction With a Decoder-Only Transformer
• TimeSynth: A Framework for Uncovering Systematic Biases in Time Series Forecasting
• Multi-Level Strategic Classification: Incentivizing Improvement through Promotion and Relegation Dynamics
• Hierarchical Concept Embedding & Pursuit for Interpretable Image Classification
• Assessing Low Back Movement with Motion Tape Sensor Data Through Deep Learning
• PRISM: A 3D Probabilistic Neural Representation for Interpretable Shape Modeling
• External Division of Two Bregman Proximity Operators for Poisson Inverse Problems
• Exploring Multiple High-Scoring Subspaces in Generative Flow Networks
• Partial GFlowNet: Accelerating Convergence in Large State Spaces via Strategic Partitioning
• A Generic Framework for Fair Consensus Clustering in Streams
• Calibrating an Imperfect Auxiliary Predictor for Unobserved No-Purchase Choice
• Unifying Stable Optimization and Reference Regularization in RLHF
• PASCAL: A Phase-Aware Scheduling Algorithm for Serving Reasoning-based Large Language Models
• Real-Time Proactive Anomaly Detection via Forward and Backward Forecast Modeling
• The Implicit Bias of Steepest Descent with Mini-batch Stochastic Gradient
• Brain4FMs: A Benchmark of Foundation Models for Electrical Brain Signal
• Learn from Your Mistakes: Self-Correcting Masked Diffusion Models
• SkillRater: Untangling Capabilities in Multimodal Data
• How Well Do Large-Scale Chemical Language Models Transfer to Downstream Tasks?
• TreeGrad-Ranker: Feature Ranking via $O(L)$-Time Gradients for Decision Trees
• GP2F: Cross-Domain Graph Prompting with Adaptive Fusion of Pre-trained Graph Neural Networks
• TIP: Resisting Gradient Inversion via Targeted Interpretable Perturbation in Federated Learning
• Both Topology and Text Matter: Revisiting LLM-guided Out-of-Distribution Detection on Text-attributed Graphs
• UMAP Is Spectral Clustering on the Fuzzy Nearest-Neighbor Graph
• Fully First-Order Algorithms for Online Bilevel Optimization
• Explainable Machine-Learning based Detection of Knee Injuries in Runners
• SpiralFormer: Looped Transformers Can Learn Hierarchical Dependencies via Multi-Resolution Recursion
• Zooming without Zooming: Region-to-Image Distillation for Fine-Grained Multimodal Perception
• Towards Fair and Comprehensive Evaluation of Routers in Collaborative LLM Systems
• SynthRAR: Ring Artifacts Reduction in CT with Unrolled Network and Synthetic Data Training
• Where Bits Matter in World Model Planning: A Paired Mixed-Bit Study for Efficient Spatial Reasoning
• Agentic AI for Cybersecurity: A Meta-Cognitive Architecture for Governable Autonomy
• Mitigating Mismatch within Reference-based Preference Optimization
• Leveraging LLMs to support co-evolution between definitions and instances of textual DSLs: A Systematic Evaluation
• DynaHOI: Benchmarking Hand-Object Interaction for Dynamic Target
• Who Does What? Archetypes of Roles Assigned to LLMs During Human-AI Decision-Making
• AdaptEvolve: Improving Efficiency of Evolutionary AI Agents through Adaptive Model Selection
• IncompeBench: A Permissively Licensed, Fine-Grained Benchmark for Music Information Retrieval
• Synthesis of Late Gadolinium Enhancement Images via Implicit Neural Representations for Cardiac Scar Segmentation
• Towards Performance-Enhanced Model-Contrastive Federated Learning using Historical Information in Heterogeneous Scenarios
• TAVAE: A VAE with Adaptable Priors Explains Contextual Modulation in the Visual Cortex
• Manifold-Aware Temporal Domain Generalization for Large Language Models
• Accelerating Robotic Reinforcement Learning with Agent Guidance
• Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?
• On the Sensitivity of Firing Rate-Based Federated Spiking Neural Networks to Differential Privacy
• An Empirical Study of the Imbalance Issue in Software Vulnerability Detection
• Fourier Transformers for Latent Crystallographic Diffusion and Generative Modeling
• ModelWisdom: An Integrated Toolkit for TLA+ Model Visualization, Digest and Repair
• Choose Your Agent: Tradeoffs in Adopting AI Advisors, Coaches, and Delegates in Multi-Party Negotiation
• DeepSight: An All-in-One LM Safety Toolkit
• Multi Graph Search for High-Dimensional Robot Motion Planning
• On the Complexity of Offline Reinforcement Learning with $Q^\star$-Approximation and Partial Coverage
• KAN-FIF: Spline-Parameterized Lightweight Physics-based Tropical Cyclone Estimation on Meteorological Satellite
• Meta-Sel: Efficient Demonstration Selection for In-Context Learning via Supervised Meta-Learning
• Learning beyond Teacher: Generalized On-Policy Distillation with Reward Extrapolation
• On the Adoption of AI Coding Agents in Open-source Android and iOS Development
• dVoting: Fast Voting for dLLMs
• 3DGSNav: Enhancing Vision-Language Model Reasoning for Object Navigation via Active 3D Gaussian Splatting
• SAGEO Arena: A Realistic Environment for Evaluating Search-Augmented Generative Engine Optimization
• Visual Reasoning Benchmark: Evaluating Multimodal LLMs on Classroom-Authentic Visual Problems from Primary Education
• DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing
• VIRENA: Virtual Arena for Research, Education, and Democratic Innovation
• The Observer Effect in World Models: Invasive Adaptation Corrupts Latent Physics
• Towards On-Policy SFT: Distribution Discriminant Theory and its Applications in LLM Training
• Bandit Learning in Matching Markets with Interviews
• Energy-Aware Spike Budgeting for Continual Learning in Spiking Neural Networks for Neuromorphic Vision
• Olmix: A Framework for Data Mixing Throughout LM Development
• Intrinsic-Energy Joint Embedding Predictive Architectures Induce Quasimetric Spaces
• ExtractBench: A Benchmark and Evaluation Methodology for Complex Structured Extraction
• A technical curriculum on language-oriented artificial intelligence in translation and specialised communication
• On the implicit regularization of Langevin dynamics with projected noise
• Creative Ownership in the Age of AI
• AttentionRetriever: Attention Layers are Secretly Long Document Retrievers
• UniT: Unified Multimodal Chain-of-Thought Test-time Scaling
• Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment
• Phase Transition for Budgeted Multi-Agent Synergy
• Compiling High-Level Neural Network Specifications into VNN-LIB Queries
• NewsInterview: a Dataset and a Playground to Evaluate LLMs' Ground Gap via Informational Interviews
• PBP: Post-training Backdoor Purification for Malware Classifiers
• SWE-MiniSandbox: Container-Free Reinforcement Learning for Building Software Engineering Agents
• Credal Concept Bottleneck Models: Structural Separation of Epistemic and Aleatoric Uncertainty
• AI-Driven Clinical Decision Support System for Enhanced Diabetes Diagnosis and Management
• Toward Reliable Tea Leaf Disease Diagnosis Using Deep Learning Model: Enhancing Robustness With Explainable AI and Adversarial Training
• How Many Features Can a Language Model Store Under the Linear Representation Hypothesis?
• DeepRed: an architecture for redshift estimation
• HiFloat4 Format for Language Model Inference
• CryptoAnalystBench: Failures in Multi-Tool Long-Form LLM Analysis
• Predictive Associative Memory: Retrieval Beyond Similarity Through Temporal Co-occurrence
• Security Threat Modeling for Emerging AI-Agent Protocols: A Comparative Analysis of MCP, A2A, Agora, and ANP
• MolmoSpaces: A Large-Scale Open Ecosystem for Robot Navigation and Manipulation
• Situated, Dynamic, and Subjective: Envisioning the Design of Theory-of-Mind-Enabled Everyday AI with Industry Practitioners
• Divide and Learn: Multi-Objective Combinatorial Optimization at Scale
• When Models Examine Themselves: Vocabulary-Activation Correspondence in Self-Referential Processing
• Bootstrapping-based Regularisation for Reducing Individual Prediction Instability in Clinical Risk Prediction Models
• Finding the Cracks: Improving LLMs Reasoning with Paraphrastic Probing and Consistency Verification
• The Energy of Falsehood: Detecting Hallucinations via Diffusion Model Likelihoods
• The Manifold of the Absolute: Religious Perennialism as Generative Inference
• Retrieval-Aware Distillation for Transformer-SSM Hybrids
• General and Efficient Steering of Unconditional Diffusion
• Can We Really Learn One Representation to Optimize All Rewards?
• When Visibility Outpaces Verification: Delayed Verification and Narrative Lock-in in Agentic AI Discourse
• Gradients Must Earn Their Influence: Unifying SFT with Generalized Entropic Objectives
• Fighting MRI Anisotropy: Learning Multiple Cardiac Shapes From a Single Implicit Neural Representation
• Towards Reliable Machine Translation: Scaling LLMs for Critical Error Detection and Safety
• Enhanced Portable Ultra Low-Field Diffusion Tensor Imaging with Bayesian Artifact Correction and Deep Learning-Based Super-Resolution
• From Noise to Order: Learning to Rank via Denoising Diffusion
• EM-Aware Physical Synthesis: Neural Inductor Modeling and Intelligent Placement & Routing for RF Circuits
• Compiler-Guided Inference-Time Adaptation: Improving GPT-5 Programming Performance in Idris
• Understanding Persuasive Interactions between Generative Social Agents and Humans: The Knowledge-based Persuasion Model (KPM)
• RooflineBench: A Benchmarking Framework for On-Device LLMs via Roofline Analysis
• Multimodal Fact-Level Attribution for Verifiable Reasoning
• Differentially Private and Communication Efficient Large Language Model Split Inference via Stochastic Quantization and Soft Prompt
• How Smart Is Your GUI Agent? A Framework for the Future of Software Interaction
• Locally Interpretable Individualized Treatment Rules for Black-Box Decision Models
• Adaptive Milestone Reward for GUI Agents
• Stop Tracking Me! Proactive Defense Against Attribute Inference Attack in LLMs
• AltTS: A Dual-Path Framework with Alternating Optimization for Multivariate Time Series Forecasting
• Krause Synchronization Transformers
• Native Reasoning Models: Training Language Models to Reason on Unverifiable Data
• TS-Memory: Plug-and-Play Memory for Time Series Foundation Models
• Perception-based Image Denoising via Generative Compression
• ReaDy-Go: Real-to-Sim Dynamic 3D Gaussian Splatting Simulation for Environment-Specific Visual Navigation with Moving Obstacles
• Analytical Search
• Gradient Compression May Hurt Generalization: A Remedy by Synthetic Data Guided Sharpness Aware Minimization
• ABot-N0: Technical Report on the VLA Foundation Model for Versatile Embodied Navigation
• PLOT-CT: Pre-log Voronoi Decomposition Assisted Generation for Low-dose CT Reconstruction
• ArGEnT: Arbitrary Geometry-encoded Transformer for Operator Learning
• ScalSelect: Scalable Training-Free Multimodal Data Selection for Efficient Visual Instruction Tuning
• Variation-aware Flexible 3D Gaussian Editing
• ViTaS: Visual Tactile Soft Fusion Contrastive Learning for Visuomotor Learning
• Brain Tumor Classifiers Under Attack: Robustness of ResNet Variants Against Transferable FGSM and PGD Attacks
• DMind-3: A Sovereign Edge--Local--Cloud AI System with Controlled Deliberation and Correction-Based Tuning for Safe, Low-Latency Transaction Execution
• LoRA-based Parameter-Efficient LLMs for Continuous Learning in Edge-based Malware Detection
• SToRM: Supervised Token Reduction for Multi-modal LLMs toward efficient end-to-end autonomous driving
• Provable Offline Reinforcement Learning for Structured Cyclic MDPs
• PatientHub: A Unified Framework for Patient Simulation
• DRACO: a Cross-Domain Benchmark for Deep Research Accuracy, Completeness, and Objectivity
• ANML: Attribution-Native Machine Learning with Guaranteed Robustness
• OMEGA-Avatar: One-shot Modeling of 360{\deg} Gaussian Avatars
• TabSieve: Explicit In-Table Evidence Selection for Tabular Prediction
• Semantically Conditioned Diffusion Models for Cerebral DSA Synthesis
• LLM-Driven 3D Scene Generation of Agricultural Simulation Environments
• Adapting Vision-Language Models for E-commerce Understanding at Scale
• AmbiBench: Benchmarking Mobile GUI Agents Beyond One-Shot Instructions in the Wild
• Cooperation Breakdown in LLM Agents Under Communication Delays
• MiniCPM-SALA: Hybridizing Sparse and Linear Attention for Efficient Long-Context Modeling
• Safe Fairness Guarantees Without Demographics in Classification: Spectral Uncertainty Set Perspective
• Evaluating LLM Safety Under Repeated Inference via Accelerated Prompt Stress Testing
• ULTRA:Urdu Language Transformer-based Recommendation Architecture
• Improving Neural Retrieval with Attribution-Guided Query Rewriting
• Resource-Aware Deployment Optimization for Collaborative Intrusion Detection in Layered Networks
• PhyNiKCE: A Neurosymbolic Agentic Framework for Autonomous Computational Fluid Dynamics
• Benchmark Health Index: A Systematic Framework for Benchmarking the Benchmarks of LLMs
• Right for the Wrong Reasons: Epistemic Regret Minimization for Causal Rung Collapse in LLMs
• Beyond Pixels: Vector-to-Graph Transformation for Reliable Schematic Auditing
• ThinkRouter: Efficient Reasoning via Routing Thinking between Latent and Discrete Spaces
• Beyond Parameter Arithmetic: Sparse Complementary Fusion for Distribution-Aware Model Merging
• Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMs
• Text2GQL-Bench: A Text to Graph Query Language Benchmark [Experiment, Analysis & Benchmark]
• AIR: Improving Agent Safety through Incident Response
• TSR: Trajectory-Search Rollouts for Multi-Turn RL of LLM Agents
• How to Optimize Multispecies Set Predictions in Presence-Absence Modeling ?
• RELATE: A Reinforcement Learning-Enhanced LLM Framework for Advertising Text Generation
• FlowMind: Execute-Summarize for Structured Workflow Generation from LLM Reasoning
• Beyond End-to-End Video Models: An LLM-Based Multi-Agent System for Educational Video Generation
• Detecting RLVR Training Data via Structural Convergence of Reasoning
• Hi-SAM: A Hierarchical Structure-Aware Multi-modal Framework for Large-Scale Recommendation
• PuYun-LDM: A Latent Diffusion Model for High-Resolution Ensemble Weather Forecasts
• Predicting LLM Output Length via Entropy-Guided Representations
• Revis: Sparse Latent Steering to Mitigate Object Hallucination in Large Vision-Language Models
• Prototype Transformer: Towards Language Model Architectures Interpretable by Design
• Talk2DM: Enabling Natural Language Querying and Commonsense Reasoning for Vehicle-Road-Cloud Integrated Dynamic Maps with Large Language Models
• Intelligent AI Delegation
• From Atoms to Trees: Building a Structured Feature Forest with Hierarchical Sparse Autoencoders
• When Should LLMs Be Less Specific? Selective Abstraction for Reliable Long-Form Text Generation
• AlphaPROBE: Alpha Mining via Principled Retrieval and On-graph biased evolution
• MEME: Modeling the Evolutionary Modes of Financial Markets
• Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments
• CSEval: A Framework for Evaluating Clinical Semantics in Text-to-Image Generation
• InjectRBP: Steering Large Language Model Reasoning Behavior via Pattern Injection
• Multi UAVs Preflight Planning in a Shared and Dynamic Airspace
• LawThinker: A Deep Research Legal Agent in Dynamic Environments
• Tiny Recursive Reasoning with Mamba-2 Attention Hybrid
• Differentiable Modal Logic for Multi-Agent Diagnosis, Orchestration and Communication
• The Pensieve Paradigm: Stateful Language Models Mastering Their Own Context
• Stop Unnecessary Reflection: Training LRMs for Efficient Reasoning with Adaptive Reflection and Length Coordinated Penalty
• Commencing-Student Enrolment Forecasting Under Data Sparsity with Time Series Foundation Models
• HLA: Hadamard Linear Attention
• Neutral Prompts, Non-Neutral People: Quantifying Gender and Skin-Tone Bias in Gemini Flash 2.5 Image and GPT Image 1.5
• Value Alignment Tax: Measuring Value Trade-offs in LLM Alignment
• STAR : Bridging Statistical and Agentic Reasoning for Large Model Performance Prediction
• Seq2Seq2Seq: Lossless Data Compression via Discrete Latent Transformers and Reinforcement Learning
• GPT-4o Lacks Core Features of Theory of Mind
• Sci-CoE: Co-evolving Scientific Reasoning LLMs via Geometric Consensus with Sparse Supervision
• Statistical Parsing for Logical Information Retrieval
• Pedagogically-Inspired Data Synthesis for Language Model Knowledge Distillation
• SAM3-LiteText: An Anatomical Study of the SAM3 Text Encoder for Efficient Vision-Language Segmentation
• "Sorry, I Didn't Catch That": How Speech Models Miss What Matters Most
• Think like a Scientist: Physics-guided LLM Agent for Equation Discovery
• CM2: Reinforcement Learning with Checklist Rewards for Multi-Turn and Multi-Step Agentic Tool Use
• Agentic Test-Time Scaling for WebAgents
• HybridRAG: A Practical LLM-based ChatBot Framework based on Pre-Generated Q&A over Raw Unstructured Documents
• Methodological Variation in Studying Staff and Student Perceptions of AI
• BIRD: A Museum Open Dataset Combining Behavior Patterns and Identity Types to Better Model Visitors' Experience
• Nested Named Entity Recognition in Plasma Physics Research Articles
• Automated Optimization Modeling via a Localizable Error-Driven Perspective
• Assessing LLM Reliability on Temporally Recent Open-Domain Questions
• Small Updates, Big Doubts: Does Parameter-Efficient Fine-tuning Enhance Hallucination Detection ?
• Visualizing and Benchmarking LLM Factual Hallucination Tendencies via Internal State Analysis and Clustering
• Enhancing SDG-Text Classification with Combinatorial Fusion Analysis and Generative AI
• Disentangling Direction and Magnitude in Transformer Representations: A Double Dissociation Through L2-Matched Perturbation Analysis
• Efficient Hyper-Parameter Search for LoRA via Language-aided Bayesian Optimization
• The Script Tax: Measuring Tokenization-Driven Efficiency and Latency Disparities in Multilingual Language Models
• Evaluating Few-Shot Temporal Reasoning of LLMs for Human Activity Prediction in Smart Environments
• What Do LLMs Know About Alzheimer's Disease? Fine-Tuning, Probing, and Data Synthesis for AD Detection
• From Instruction to Output: The Role of Prompting in Modern NLG
• KBVQ-MoE: KLT-guided SVD with Bias-Corrected Vector Quantization for MoE Large Language Models
• Spectra: Rethinking Optimizers for LLMs Under Spectral Anisotropy
• TDPNavigator-Placer: Thermal- and Wirelength-Aware Chiplet Placement in 2.5D Systems Through Multi-Agent Reinforcement Learning
• MuCO: Generative Peptide Cyclization Empowered by Multi-stage Conformation Optimization
• Time-TK: A Multi-Offset Temporal Interaction Framework Combining Transformer and Kolmogorov-Arnold Networks for Time Series Forecasting
• MELINOE: Fine-Tuning Enables Memory-Efficient Inference for Mixture-of-Experts Models
• Position-Aware Self-supervised Representation Learning for Cross-mode Radar Signal Recognition
• Hybrid operator learning of wave scattering maps in high-contrast media
• DDL2PropBank Agent: Benchmarking Multi-Agent Frameworks' Developer Experience Through a Novel Relational Schema Mapping Task
• interwhen: A Generalizable Framework for Verifiable Reasoning with Test-time Monitors
• Zero-Sacrifice Persistent-Robustness Adversarial Defense for Pre-Trained Encoders
• UltraLIF: Fully Differentiable Spiking Neural Networks via Ultradiscretization and Max-Plus Algebra
• Explaining AI Without Code: A User Study on Explainable AI
• Latent Generative Solvers for Generalizable Long-Term Physics Simulation
• On Decision-Valued Maps and Representational Dependence
• Voxtral Realtime
• The PBSAI Governance Ecosystem: A Multi-Agent AI Reference Architecture for Securing Enterprise AI Estates
• Dissecting Subjectivity and the "Ground Truth" Illusion in Data Annotation
• Bi-Level Prompt Optimization for Multimodal LLM-as-a-Judge
• AgentNoiseBench: Benchmarking Robustness of Tool-Using LLM Agents Under Noisy Condition
• Pushing Forward Pareto Frontiers of Proactive Agents with Behavioral Agentic Optimization
• ReplicatorBench: Benchmarking LLM Agents for Replicability in Social and Behavioral Sciences
• Causal-JEPA: Learning World Models through Object-Level Latent Interventions
• GHOST: Unmasking Phantom States in Mamba2 via Grouped Hidden-state Output-aware Selection & Truncation
• TRACER: Trajectory Risk Aggregation for Critical Episodes in Agentic Reasoning
• Distributionally Robust Cooperative Multi-Agent Reinforcement Learning via Robust Value Factorization
• Credit Where It is Due: Cross-Modality Connectivity Drives Precise Reinforcement Learning for MLLM Reasoning
• AgentLeak: A Full-Stack Benchmark for Privacy Leakage in Multi-Agent LLM Systems
• Human-Inspired Continuous Learning of Internal Reasoning Processes: Learning How to Think for Adaptive AI Systems
• CausalAgent: A Conversational Multi-Agent System for End-to-End Causal Inference
• Budget-Constrained Agentic Large Language Models: Intention-Based Planning for Costly Tool Use
• SemaPop: Semantic-Persona Conditioned Population Synthesis
• Learning to Configure Agentic AI Systems
• The Five Ws of Multi-Agent Communication: Who Talks to Whom, When, What, and Why -- A Survey from MARL to Emergent Language and LLMs
• MAPLE: Modality-Aware Post-training and Learning Ecosystem
• scPilot: Large Language Model Reasoning Toward Automated Single-Cell Analysis and Discovery
• When Agents Disagree With Themselves: Measuring Behavioral Consistency in LLM-Based Agents
• Neuro-Symbolic Multitasking: A Unified Framework for Discovering Generalizable Solutions to PDE Families
• Do MLLMs Really Understand Space? A Mathematical Reasoning Evaluation
• Quark Medical Alignment: A Holistic Multi-Dimensional Alignment and Collaborative Optimization Paradigm