AI Research News Feeds for October 22nd, 2025

AI RESEARCH PAPERS & ACADEMIC SOURCES

• WHAT-IF: Exploring Branching Narratives by Meta-Prompting Large Language Models
• DialUp! Modeling the Language Continuum by Adapting Models to Dialects and Dialects to Models
• DCAD-2000: A Multilingual Dataset across 2000+ Languages with Data Cleaning as Anomaly Detection
• Think Silently, Think Fast: Dynamic Latent Compression of LLM Reasoning Chains
• CoDial: Interpretable Task-Oriented Dialogue Systems Through Dialogue Flow Alignment
• The Translation Barrier Hypothesis: Multilingual Generation with Large Language Models Suffers from Implicit Translation Failure
• Fairshare Data Pricing via Data Valuation for Large Language Models
• DrunkAgent: Stealthy Memory Corruption in LLM-Powered Recommender Agents
• Don't Retrieve, Generate: Prompting LLMs for Synthetic Training Data in Dense Retrieval
• Facts are Harder Than Opinions -- A Multilingual, Comparative Analysis of LLM-Based Fact-Checking Reliability
• MSR-Align: Policy-Grounded Multimodal Alignment for Safety-Aware Reasoning in Vision-Language Models
• CoIDO: Efficient Data Selection for Visual Instruction Tuning via Coupled Importance-Diversity Optimization
• CMIS-Net: A Cascaded Multi-Scale Individual Standardization Network for Backchannel Agreement Estimation
• Robotic Classification of Divers' Swimming States using Visual Pose Keypoints as IMUs
• InsideOut: Integrated RGB-Radiative Gaussian Splatting for Comprehensive 3D Object Representation
• GAN-based Content-Conditioned Generation of Handwritten Musical Symbols
• Investigating Demographic Bias in Brain MRI Segmentation: A Comparative Study of Deep-Learning and Non-Deep-Learning Methods
• ManzaiSet: A Multimodal Dataset of Viewer Responses to Japanese Manzai Comedy
• Chimera: Compositional Image Generation using Part-based Concepting
• Big Data, Tiny Targets: An Exploratory Study in Machine Learning-enhanced Detection of Microplastic from Filters
• From Volume Rendering to 3D Gaussian Splatting: Theory and Applications
• Online In-Context Distillation for Low-Resource Vision Language Models
• World-in-World: World Models in a Closed-Loop World
• Adapting Stereo Vision From Objects To 3D Lunar Surface Reconstruction with the StereoLunar Dataset
• EMA-SAM: Exponential Moving-average for SAM-based PTMC Segmentation
• Beyond Frequency: Scoring-Driven Debiasing for Object Detection via Blueprint-Prompted Image Synthesis
• DeepSeek-OCR: Contexts Optical Compression
• BlendCLIP: Bridging Synthetic and Real Domains for Zero-Shot 3D Object Classification with Multimodal Pretraining
• OpenInsGaussian: Open-vocabulary Instance Gaussian Segmentation with Context-aware Cross-view Fusion
• UWBench: A Comprehensive Vision-Language Benchmark for Underwater Understanding
• TreeFedDG: Alleviating Global Drift in Federated Domain Generalization for Medical Image Segmentation
• GeoDiff: Geometry-Guided Diffusion for Metric Depth Estimation
• Proactive Reasoning-with-Retrieval Framework for Medical Multimodal Large Language Models
• OmniNWM: Omniscient Driving Navigation World Models
• Beyond Single Models: Mitigating Multimodal Hallucinations via Adaptive Token Ensemble Decoding
• Enhancing Few-Shot Classification of Benchmark and Disaster Imagery with ATTBHFA-Net
• ViSE: A Systematic Approach to Vision-Only Street-View Extrapolation
• GPTFace: Generative Pre-training of Facial-Linguistic Transformer by Span Masking and Weakly Correlated Text-image Data
• AV-Master: Dual-Path Comprehensive Perception Makes Better Audio-Visual Question Answering
• Ranking-based Preference Optimization for Diffusion Models from Implicit User Feedback
• Learning Human-Object Interaction as Groups
• FeatureFool: Zero-Query Fooling of Video Models via Feature Map
• Cross-Modal Scene Semantic Alignment for Image Complexity Assessment
• Entropy-Enhanced Conformal Features from Ricci Flow for Robust Alzheimer's Disease Classification
• Bayesian Fully-Connected Tensor Network for Hyperspectral-Multispectral Image Fusion
• Beyond Single Images: Retrieval Self-Augmented Unsupervised Camouflaged Object Detection
• LAND: Lung and Nodule Diffusion for 3D Chest CT Synthesis with Anatomical Guidance
• Mono4DGS-HDR: High Dynamic Range 4D Gaussian Splatting from Alternating-exposure Monocular Videos
• DWaste: Greener AI for Waste Sorting using Mobile and Edge Devices
• RayPose: Ray Bundling Diffusion for Template Views in Unseen 6D Object Pose Estimation
• GBlobs: Local LiDAR Geometry for Improved Sensor Placement Generalization
• Descriptor: Occluded nuScenes: A Multi-Sensor Dataset for Evaluating Perception Robustness in Automated Driving
• Image augmentation with invertible networks in interactive satellite image change detection
• Beyond the Pipeline: Analyzing Key Factors in End-to-End Deep Learning for Historical Writer Identification
• MoGA: Mixture-of-Groups Attention for End-to-End Long Video Generation
• UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generation
• Exploring a Unified Vision-Centric Contrastive Alternatives on Multi-Modal Web Documents
• A Renaissance of Explicit Motion Information Mining from Transformers for Action Recognition
• PLANA3R: Zero-shot Metric Planar 3D Reconstruction via Feed-Forward Planar Splatting
• SSD: Spatial-Semantic Head Decoupling for Efficient Autoregressive Image Generation
• IF-VidCap: Can Video Caption Models Follow Instructions?
• Moving Light Adaptive Colonoscopy Reconstruction via Illumination-Attenuation-Aware 3D Gaussian Splatting
• SEAL: Semantic-Aware Hierarchical Learning for Generalized Category Discovery
• Detection and Simulation of Urban Heat Islands Using a Fine-Tuned Geospatial Foundation Model for Microclimate Impact Prediction
• UltraGen: High-Resolution Video Generation with Hierarchical Attention
• Rebellious Student: A Complementary Learning Framework for Background Feature Enhancement in Hyperspectral Anomaly Detection
• ProCLIP: Progressive Vision-Language Alignment via LLM-based Embedder
• A Geometric Approach to Steerable Convolutions
• SAM 2++: Tracking Anything at Any Granularity
• Unifying and Enhancing Graph Transformers via a Hierarchical Mask Framework
• FedDEAP: Adaptive Dual-Prompt Tuning for Multi-Domain Federated Learning
• DSI-Bench: A Benchmark for Dynamic Spatial Intelligence
• Robobench: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models as Embodied Brain
• Cross-Domain Multi-Person Human Activity Recognition via Near-Field Wi-Fi Sensing
• DMTrack: Deformable State-Space Modeling for UAV Multi-Object Tracking with Kalman Fusion and Uncertainty-Aware Association
• Conformal Lesion Segmentation for 3D Medical Images
• A Generalizable Light Transport 3D Embedding for Global Illumination
• DualHash: A Stochastic Primal-Dual Algorithm with Theoretical Guarantee for Deep Hashing
• CUARewardBench: A Benchmark for Evaluating Reward Models on Computer-using Agent
• When LLMs step into the 3D World: A Survey and Meta-Analysis of 3D Tasks via Multi-modal Large Language Models
• Learning Collaborative Knowledge with Multimodal Representation for Polyp Re-Identification
• H3D-DGS: Exploring Heterogeneous 3D Motion Representation for Deformable 3D Gaussian Splatting
• 3D Audio-Visual Segmentation
• Med-2E3: A 2D-Enhanced 3D Medical Multimodal Large Language Model
• Foundation Cures Personalization: Improving Personalized Models' Prompt Consistency via Hidden Foundation Knowledge
• View Transformation Robustness for Multi-View 3D Object Reconstruction with Reconstruction Error-Guided View Selection
• WMamba: Wavelet-based Mamba for Face Forgery Detection
• ITVTON: Virtual Try-On Diffusion Transformer Based on Integrated Image and Text
• RAD: Training an End-to-End Driving Policy via Large-Scale 3DGS-based Reinforcement Learning
• Mask Image Watermarking
• Monitoring morphometric drift in lifelong learning segmentation of the spinal cord
• VisualQuality-R1: Reasoning-Induced Image Quality Assessment via Reinforcement Learning to Rank
• Visionary-R1: Mitigating Shortcuts in Visual Reasoning with Reinforcement Learning
• A Unified Solution to Video Fusion: From Multi-Frame Learning to Benchmarking
• DisasterM3: A Remote Sensing Vision-Language Dataset for Disaster Damage Assessment and Response
• Re-ttention: Ultra Sparse Visual Generation via Attention Statistical Reshape
• Janus-Pro-R1: Advancing Collaborative Visual Comprehension and Generation via Reinforcement Learning
• From Objects to Anywhere: A Holistic Benchmark for Multi-level Visual Grounding in 3D Scenes
• SDTagNet: Leveraging Text-Annotated Navigation Maps for Online HD Map Construction
• ReID5o: Achieving Omni Multi-modal Person Re-identification in a Single Model
• HOIDiNi: Human-Object Interaction through Diffusion Noise Optimization
• Polyline Path Masked Attention for Vision Transformer
• GreenHyperSpectra: A multi-source hyperspectral dataset for global vegetation trait prediction
• RODS: Robust Optimization Inspired Diffusion Sampling for Detecting and Reducing Hallucination in Generative Models
• Moving Object Detection from Moving Camera Using Focus of Expansion Likelihood and Segmentation
• RWKV-UNet: Improving UNet with Long-Range Cooperation for Effective Medical Image Segmentation
• SimCortex: Collision-free Simultaneous Cortical Surfaces Reconstruction
• ViBED-Net: Video Based Engagement Detection Network Using Face-Aware and Scene-Aware Spatiotemporal Cues
• Transformer Redesign for Late Fusion of Audio-Text Features on Ultra-Low-Power Edge Hardware
• Fast Agnostic Learners in the Plane
• Arbitrated Indirect Treatment Comparisons
• PrivaDE: Privacy-preserving Data Evaluation for Blockchain-based Data Marketplaces
• Generalization Below the Edge of Stability: The Role of Data Geometry
• Extracting Rule-based Descriptions of Attention Features in Transformers
• Beating the Winner's Curse via Inference-Aware Policy Optimization
• Joint Estimation of Piano Dynamics and Metrical Structure with a Multi-task Multi-Scale Network
• RESCUE: Retrieval Augmented Secure Code Generation
• The Bias-Variance Tradeoff in Data-Driven Optimization: A Local Misspecification Perspective
• LIME: Link-based user-item Interaction Modeling with decoupled xor attention for Efficient test time scaling
• Efficient Few-shot Identity Preserving Attribute Editing for 3D-aware Deep Generative Models
• A Distributed Framework for Causal Modeling of Performance Variability in GPU Traces
• Parametrising the Inhomogeneity Inducing Capacity of a Training Set, and its Impact on Supervised Learning
• ECG-LLM-- training and evaluation of domain-specific large language models for electrocardiography
• A machine learning approach to automation and uncertainty evaluation for self-validating thermocouples
• Vision Foundation Models Can Be Good Tokenizers for Latent Diffusion Models
• Decoding Dynamic Visual Experience from Calcium Imaging via Cell-Pattern-Aware SSL
• Interval Prediction of Annual Average Daily Traffic on Local Roads via Quantile Random Forest with High-Dimensional Spatial Data
• A Multi-Evidence Framework Rescues Low- Power Prognostic Signals and Rejects Statistical Artifacts in Cancer Genomics
• CovMatch: Cross-Covariance Guided Multimodal Dataset Distillation with Trainable Text Encoder
• Channel-Aware Vector Quantization for Robust Semantic Communication on Discrete Channels
• A Compositional Paradigm for Foundation Models: Towards Smarter Robotic Agents
• Differentially Private E-Values
• Bayesian Low-Rank Factorization for Robust Model Adaptation
• Adapting Language Balance in Code-Switching Speech
• Diffusion Buffer for Online Generative Speech Enhancement
• Symbolic Emulators for Cosmology: Accelerating Cosmological Analyses Without Sacrificing Precision
• Analyse comparative d'algorithmes de restauration en architecture d\'epli\'ee pour des signaux chromatographiques parcimonieux
• A Frequentist Statistical Introduction to Variational Inference, Autoencoders, and Diffusion Models
• SO(3)-invariant PCA with application to molecular data
• MTraining: Distributed Dynamic Sparse Attention for Efficient Ultra-Long Context Training
• One-Pass Learning via Bridging Orthogonal Gradient Descent and Recursive Least-Squares
• Better NTK Conditioning: A Free Lunch from (ReLU) Nonlinear Activation in Wide Neural Networks
• Sparse Explanations of Neural Networks Using Pruned Layer-Wise Relevance Propagation
• A Flow-Based Model for Conditional and Probabilistic Electricity Consumption Profile Generation and Prediction
• Learning Confidence Bounds for Classification with Imbalanced Data
• One protein is all you need
• FedMeld: A Model-dispersal Federated Learning Framework for Space-ground Integrated Networks
• Asynchronous Federated Learning: A Scalable Approach for Decentralized Machine Learning
• A Statistical Theory of Contrastive Pre-training and Multimodal Generative AI
• Can We Validate Counterfactual Estimations in the Presence of General Network Interference?
• Analyzing Similarity Metrics for Data Selection for Language Model Pretraining
• Beyond Benign Overfitting in Nadaraya-Watson Interpolators
• In-Context Learning of Linear Dynamical Systems with Transformers: Approximation Bounds and Depth-Separation
• Enhancing Sample Selection Against Label Noise by Cutting Mislabeled Easy Examples
• CayleyPy RL: Pathfinding and Reinforcement Learning on Cayley Graphs
• In-Context Learning of Stochastic Differential Equations with Foundation Inference Models
• Reinforcement Learning with Verifiable Rewards: GRPO's Effective Loss, Dynamics, and Success Amplification
• T\'yr-the-Pruner: Structural Pruning LLMs via Global Sparsity Distribution Optimization
• PAUSE: Low-Latency and Privacy-Aware Active User Selection for Federated Learning
• Efficient Verified Machine Unlearning For Distillation
• Enabling Automatic Differentiation with Mollified Graph Neural Operators
• Spike-timing-dependent Hebbian learning as noisy gradient descent
• FlashBias: Fast Computation of Attention with Bias
• Neural Graduated Assignment for Maximum Common Edge Subgraphs
• Fair Supervised Learning Through Constraints on Smooth Nonconvex Unfairness-Measure Surrogates
• Understanding Differential Transformer Unchains Pretrained Self-Attentions
• The Spacetime of Diffusion Models: An Information Geometry Perspective
• Inverse Q-Learning Done Right: Offline Imitation Learning in $Q^\pi$-Realizable MDPs
• Generative or Discriminative? Revisiting Text Classification in the Era of Transformers
• PARALLELPROMPT: Extracting Parallelism from Large Language Model Queries
• A unified framework for establishing the universal approximation of transformer-type architectures
• Class-wise Balancing Data Replay for Federated Class-Incremental Learning
• The Impact of Coreset Selection on Spurious Correlations and Group Robustness
• Graph Neural Networks for Road Safety Modeling: Datasets and Evaluations for Accident Analysis
• Generation of Uncertainty-Aware Emergent Concepts in Factorized 3D Scene Graphs via Graph Neural Networks
• Implicit Neural Compression of Point Clouds
• On Generalization and Distributional Update for Mimicking Observations with Adequate Exploration
• Molecular Fingerprints Are Strong Models for Peptide Function Prediction
• Dynamic object goal pushing with mobile manipulators through model-free constrained reinforcement learning
• The $\varphi$ Curve: The Shape of Generalization through the Lens of Norm-based Capacity Control
• VLA-Cache: Efficient Vision-Language-Action Manipulation via Adaptive Token Caching
• Improving Diffusion-based Inverse Algorithms under Few-Step Constraint via Learnable Linear Extrapolation
• Harnessing Test-time Adaptation for NLU tasks Involving Dialects of English
• Low-cost Embedded Breathing Rate Determination Using 802.15.4z IR-UWB Hardware for Remote Healthcare
• REPA-E: Unlocking VAE for End-to-End Tuning with Latent Diffusion Transformers
• A Physics-Informed Spatiotemporal Deep Learning Framework for Turbulent Systems
• Backward Conformal Prediction
• Time Reversal Symmetry for Efficient Robotic Manipulations in Deep Reinforcement Learning
• Steering Generative Models with Experimental Data for Protein Fitness Optimization
• gen2seg: Generative Models Enable Generalizable Instance Segmentation
• DanmakuTPPBench: A Multi-modal Benchmark for Temporal Point Process Modeling and Understanding
• Nearly Dimension-Independent Convergence of Mean-Field Black-Box Variational Inference
• FlySearch: Exploring how vision-language models explore
• ALINE: Joint Amortization for Bayesian Inference and Active Data Acquisition
• Dynamic Diffusion Schr\"odinger Bridge in Astrophysical Observational Inversions
• ReVeal: Self-Evolving Code Agents via Reliable Self-Verification
• Advances in Pre-trained Language Models for Domain-Specific Text Classification: A Systematic Review
• Atomic Literary Styling: Mechanistic Manipulation of Prose Generation in Neural Language Models
• Chain-of-Thought Reasoning Improves Context-Aware Translation with Large Language Models
• Na Pr\'atica, qual IA Entende o Direito? Um Estudo Experimental com IAs Generalistas e uma IA Jur\'idica
• Does Reasoning Help LLM Agents Play Dungeons and Dragons? A Prompt Engineering Experiment
• LLMs Encode How Difficult Problems Are
• CMT-Bench: Cricket Multi-Table Generation Benchmark for Probing Robustness in Large Language Models
• MARCUS: An Event-Centric NLP Pipeline that generates Character Arcs from Narratives
• BrailleLLM: Braille Instruction Tuning with Large Language Models for Braille Domain Tasks
• Food4All: A Multi-Agent Framework for Real-time Free Food Discovery with Integrated Nutritional Metadata
• Combining Distantly Supervised Models with In Context Learning for Monolingual and Cross-Lingual Relation Extraction
• KrishokBondhu: A Retrieval-Augmented Voice-Based Agricultural Advisory Call Center for Bengali Farmers
• KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs
• Towards Fair ASR For Second Language Speakers Using Fairness Prompted Finetuning
• Adamas: Hadamard Sparse Attention for Efficient Long-Context Inference
• Chain-of-Conceptual-Thought: Eliciting the Agent to Deeply Think within the Response
• Grounding or Guessing? Visual Signals for Detecting Hallucinations in Sign Language Translation
• Engagement Undermines Safety: How Stereotypes and Toxicity Shape Humor in Language Models
• ChronoPlay: A Framework for Modeling Dual Dynamics and Authenticity in Game RAG Benchmarks
• DePass: Unified Feature Attributing by Simple Decomposed Forward Pass
• CEFR-Annotated WordNet: LLM-Based Proficiency-Guided Semantic Database for Language Learning
• IMB: An Italian Medical Benchmark for Question Answering
• DART: A Structured Dataset of Regulatory Drug Documents in Italian for Clinical NLP
• How Efficient Are Diffusion Language Models? A Critical Examination of Efficiency Evaluation Practices
• Identity-Aware Large Language Models require Cultural Reasoning
• Building Trust in Clinical LLMs: Bias Analysis and Dataset Transparency
• Beyond the Explicit: A Bilingual Dataset for Dehumanization Detection in Social Media
• Dynamical model parameters from ultrasound tongue kinematics
• MLMA: Towards Multilingual with Mamba Based Architectures
• Investigating LLM Capabilities on Long Context Comprehension for Medical Question Answering
• SemiAdapt and SemiLoRA: Efficient Domain Adaptation for Transformer-based Low-Resource Language Translation with a Case Study on Irish
• Topoformer: brain-like topographic organization in Transformer language models through spatial querying and reweighting
• AI use in American newspapers is widespread, uneven, and rarely disclosed
• KAT-Coder Technical Report
• WebSeer: Training Deeper Search Agents through Reinforcement Learning with Self-Reflection
• HouseTour: A Virtual Real Estate A(I)gent
• The Impact of Image Resolution on Biomedical Multimodal Large Language Models
• Position: LLM Watermarking Should Align Stakeholders' Incentives for Practical Adoption
• See the Text: From Tokenization to Visual Reading
• Unconditional Truthfulness: Learning Unconditional Uncertainty of Large Language Models
• Rethinking LLM Uncertainty: A Multi-Agent Approach to Estimating Black-Box Model Uncertainty
• Discovering the curriculum with AI: A proof-of-concept demonstration with an intelligent tutoring system for teaching project selection
• LENS: Large Pre-trained Transformer for Exploring Financial Time Series Regularities
• Counterfactual Effect Decomposition in Multi-Agent Sequential Decision Making
• InternLM2.5-StepProver: Advancing Automated Theorem Proving via Critic-Guided Search
• Do LLMs Strategically Reveal, Conceal, and Infer Information? A Theoretical and Empirical Analysis in The Chameleon Game
• Modeling Human Beliefs about AI Behavior for Scalable Oversight
• A representational framework for learning and encoding structurally enriched trajectories in complex agent environments
• HyperGraphRAG: Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation
• Improving Human-AI Coordination through Online Adversarial Training and Generative Models
• Pretraining a Shared Q-Network for Data-Efficient Offline Reinforcement Learning
• MTRE: Multi-Token Reliability Estimation for Hallucination Detection in VLMs
• SOCIA: Joint Structure-Parameter Co-Optimization for Automated Simulator Construction
• Can Agents Fix Agent Issues?
• VIKI-R: Coordinating Embodied Multi-Agent Cooperation via Reinforcement Learning
• Mitigating Prior Errors in Causal Structure Learning: A Resilient Approach via Bayesian Networks
• Learning by Watching: A Review of Video-based Learning Approaches for Robot Manipulation
• Exploring Data-Efficient Adaptation of Large Language Models for Code Generation
• A Survey of Automatic Hallucination Evaluation on Natural Language Generation
• Learning Fairer Representations with FairVIC
• Review of Explainable Graph-Based Recommender Systems
• BlockScan: Detecting Anomalies in Blockchain Transactions
• Transition of $\alpha$-mixing in Random Iterations with Applications in Queuing Theory
• When Text Embedding Meets Large Language Model: A Comprehensive Survey
• Deep Learning in Palmprint Recognition-A Comprehensive Survey
• LLM Safety Alignment is Divergence Estimation in Disguise
• Foundations of a Developmental Design Paradigm for Integrated Continual Learning, Deliberative Behavior, and Comprehensibility
• Challenges in Testing Large Language Model Based Software: A Faceted Taxonomy
• Temporal Alignment of LLMs through Cycle Encoding for Long-Range Time Representations
• Energy Matching: Unifying Flow Matching and Energy-Based Models for Generative Modeling
• VLLFL: A Vision-Language Model Based Lightweight Federated Learning Framework for Smart Agriculture
• Dendritic Computing with Multi-Gate Ferroelectric Field-Effect Transistors
• Regression is all you need for medical image translation
• VITA-Audio: Fast Interleaved Cross-Modal Token Generation for Efficient Large Speech-Language Model
• The Shift Towards Preprints in AI Policy Research: A Comparative Study of Preprint Trends in the U.S., Europe, and South Korea
• MetaBox-v2: A Unified Benchmark Platform for Meta-Black-Box Optimization
• COLORA: Efficient Fine-Tuning for Convolutional Models with a Study Case on Optical Coherence Tomography Image Classification
• GraSS: Scalable Data Attribution with Gradient Sparsification and Sparse Projection
• LIMOPro: Reasoning Refinement for Efficient and Effective Test-time Scaling
• Multi-Agent Collaboration via Evolving Orchestration
• Segment Policy Optimization: Effective Segment-Level Credit Assignment in RL for Large Language Models
• REOrdering Patches Improves Vision Models
• EvaLearn: Quantifying the Learning Capability and Efficiency of LLMs via Sequential Problem Solving
• Counterfactual reasoning: an analysis of in-context emergence
• Model-based Implicit Neural Representation for sub-wavelength Radio Localization
• Mind the Web: The Security of Web Use Agents
• Understanding In-Context Learning on Structured Manifolds: Bridging Attention to Kernel Methods
• C-SEO Bench: Does Conversational SEO Work?
• Sparse Feature Coactivation Reveals Causal Semantic Modules in Large Language Models
• Iterative Quantum Feature Maps
• UniPixel: Unified Object Referring and Segmentation for Pixel-Level Visual Reasoning
• From Noise to Laws: Regularized Time-Series Forecasting via Denoised Dynamic Graphs
• Shock-Aware Physics-Guided Fusion-DeepONet Operator for Rarefied Micro-Nozzle Flows
• Demystifying Transition Matching: When and Why It Can Beat Flow Matching
• Attention-Guided Deep Adversarial Temporal Subspace Clustering (A-DATSC) Model for multivariate spatiotemporal data
• Benchmarking Probabilistic Time Series Forecasting Models on Neural Activity
• Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods
• MEG-GPT: A transformer-based foundation model for magnetoencephalography data
• Provably Optimal Reinforcement Learning under Safety Filtering
• Gradient Variance Reveals Failure Modes in Flow-Based Generative Models
• Efficient Long-context Language Model Training by Core Attention Disaggregation
• HyperDiffusionFields (HyDiF): Diffusion-Guided Hypernetworks for Learning Implicit Molecular Neural Fields
• Rethinking PCA Through Duality
• Nash Policy Gradient: A Policy Gradient Method with Iteratively Refined Regularization for Finding Nash Equilibria
• Ensemble based Closed-Loop Optimal Control using Physics-Informed Neural Networks
• Joint Optimization of Cooperation Efficiency and Communication Covertness for Target Detection with AUVs
• Towards Fast LLM Fine-tuning through Zeroth-Order Optimization with Projected Gradient-Aligned Perturbations
• ACTG-ARL: Differentially Private Conditional Text Generation with RL-Boosted Control
• Fostering the Ecosystem of AI for Social Impact Requires Expanding and Strengthening Evaluation Standards
• Learning with Dual-level Noisy Correspondence for Multi-modal Entity Alignment
• From Competition to Synergy: Unlocking Reinforcement Learning for Subject-Driven Image Generation
• Online Time Series Forecasting with Theoretical Guarantees
• Physics-Informed Parametric Bandits for Beam Alignment in mmWave Communications
• Towards Identifiability of Hierarchical Temporal Causal Representation Learning
• Uncertainty Estimation by Flexible Evidential Deep Learning
• Why Policy Gradient Algorithms Work for Undiscounted Total-Reward MDPs
• Computable universal online learning
• Ensembling Pruned Attention Heads For Uncertainty-Aware Efficient Transformers
• Learning to Flow from Generative Pretext Tasks for Neural Architecture Encoding
• Towards Unsupervised Open-Set Graph Domain Adaptation via Dual Reprogramming
• Training Diverse Graph Experts for Ensembles: A Systematic Empirical Study
• Approximation Rates of Shallow Neural Networks: Barron Spaces, Activation Functions and Optimality Analysis
• Provable Generalization Bounds for Deep Neural Networks with Adaptive Regularization
• Learning Boltzmann Generators via Constrained Mass Transport
• Safe But Not Sorry: Reducing Over-Conservatism in Safety Critics via Uncertainty-Aware Modulation
• Learning to Navigate Under Imperfect Perception: Conformalised Segmentation for Safe Reinforcement Learning
• Alibaba International E-commerce Product Search Competition DILAB Team Technical Report
• Partial VOROS: A Cost-aware Performance Metric for Binary Classifiers with Precision and Capacity Constraints
• HeFS: Helper-Enhanced Feature Selection via Pareto-Optimized Genetic Search
• Robustness Verification of Graph Neural Networks Via Lightweight Satisfiability Testing
• Unrolled-SINDy: A Stable Explicit Method for Non linear PDE Discovery from Sparsely Sampled Data
• Hardness of Learning Regular Languages in the Next Symbol Prediction Setting
• Optimality and NP-Hardness of Transformers in Learning Markovian Dynamical Functions
• Informed Learning for Estimating Drought Stress at Fine-Scale Resolution Enables Accurate Yield Prediction
• Learning Time-Varying Turn-Taking Behavior in Group Conversations
• Prototyping an End-to-End Multi-Modal Tiny-CNN for Cardiovascular Sensor Patches
• Learning Task-Agnostic Representations through Multi-Teacher Distillation
• Reinforcement Learning with Imperfect Transition Predictions: A Bellman-Jensen Approach
• OmniCast: A Masked Latent Diffusion Model for Weather Forecasting Across Time Scales
• Improving the Generation and Evaluation of Synthetic Data for Downstream Medical Causal Inference
• Enhancing Fractional Gradient Descent with Learned Optimizers
• CAGE: Curvature-Aware Gradient Estimation For Accurate Quantization-Aware Training
• Stick-Breaking Embedded Topic Model with Continuous Optimal Transport for Online Analysis of Document Streams
• On Biologically Plausible Learning in Continuous Time
• When LRP Diverges from Leave-One-Out in Transformers
• A Unified Perspective on Optimization in Machine Learning and Neuroscience: From Gradient Descent to Neural Adaptation
• Search Self-play: Pushing the Frontier of Agent Capability without Supervision
• BO4Mob: Bayesian Optimization Benchmarks for High-Dimensional Urban Mobility Problem
• A Hybrid Enumeration Framework for Optimal Counterfactual Generation in Post-Acute COVID-19 Heart Failure
• Retaining by Doing: The Role of On-Policy Data in Mitigating Forgetting
• In-Process Monitoring of Gear Power Honing Using Vibration Signal Analysis and Machine Learning
• Exploring Complexity Changes in Diseased ECG Signals for Enhanced Classification
• Single-Snapshot Gridless 2D-DoA Estimation for UCAs: A Joint Optimization Approach
• CLARAE: Clarity Preserving Reconstruction AutoEncoder for Denoising and Rhythm Classification of Intracardiac Electrograms
• Covariance Matrix Construction with Preprocessing-Based Spatial Sampling for Robust Adaptive Beamforming
• Neural networks for neurocomputing circuits: a computational study of tolerance to noise and activation function non-uniformity when machine learning materials properties
• Provenance of AI-Generated Images: A Vector Similarity and Blockchain-based Approach
• Shortcutting Pre-trained Flow Matching Diffusion Models is Almost Free Lunch
• Mixed Monotonicity Reachability Analysis of Neural ODE: A Trade-Off Between Tightness and Efficiency
• Three-dimensional inversion of gravity data using implicit neural representations
• Graphical model for tensor factorization by sparse sampling
• TritonRL: Training LLMs to Think and Code Triton Without Cheating
• Learning Time-Varying Graphs from Incomplete Graph Signals
• QINNs: Quantum-Informed Neural Networks
• Does GenAI Rewrite How We Write? An Empirical Study on Two-Million Preprints
• From Flows to Words: Can Zero-/Few-Shot LLMs Detect Network Intrusions? A Grammar-Constrained, Calibrated Evaluation on UNSW-NB15
• When Intelligence Fails: An Empirical Study on Why LLMs Struggle with Password Cracking
• Metrics and evaluations for computational and sustainable AI efficiency
• Hey Pentti, We Did It!: A Fully Vector-Symbolic Lisp
• MIN-Merging: Merge the Important Neurons for Model Merging
• Hierarchical Federated Unlearning for Large Language Models
• Long-Context Attention Benchmark: From Kernel Efficiency to Distributed Context Parallelism
• L-MoE: End-to-End Training of a Lightweight Mixture of Low-Rank Adaptation Experts
• Automated Algorithm Design for Auto-Tuning Optimizers
• Are LLMs Court-Ready? Evaluating Frontier Models on Indian Legal Reasoning
• The Sherpa.ai Blind Vertical Federated Learning Paradigm to Minimize the Number of Communications
• BreakFun: Jailbreaking LLMs via Schema Exploitation
• Interpretability Framework for LLMs in Undergraduate Calculus
• TACLA: An LLM-Based Multi-Agent Tool for Transactional Analysis Training in Education
• NeuCo-Bench: A Novel Benchmark Framework for Neural Embeddings in Earth Observation
• Uncertainty-Aware Post-Hoc Calibration: Mitigating Confidently Incorrect Predictions Beyond Calibration Metrics
• Self-Evidencing Through Hierarchical Gradient Decomposition: A Dissipative System That Maintains Non-Equilibrium Steady-State by Minimizing Variational Free Energy
• Data Unlearning Beyond Uniform Forgetting via Diffusion Time and Frequency Selection
• JT-Safe: Intrinsically Enhancing the Safety and Trustworthiness of LLMs
• ParaVul: A Parallel Large Language Model and Retrieval-Augmented Framework for Smart Contract Vulnerability Detection
• CBINNS: Cancer Biology-Informed Neural Network for Unknown Parameter Estimation and Missing Physics Identification
• CLAWS:Creativity detection for LLM-generated solutions using Attention Window of Sections
• Select-Then-Decompose: From Empirical Analysis to Adaptive Selection Strategy for Task Decomposition in Large Language Models
• Rewarding the Journey, Not Just the Destination: A Composite Path and Answer Self-Scoring Reward Mechanism for Test-Time Reinforcement Learning
• Efficient Toxicity Detection in Gaming Chats: A Comparative Study of Embeddings, Fine-Tuned Transformers and LLMs
• SpecAgent: A Speculative Retrieval and Forecasting Agent for Code Completion
• EvoSyn: Generalizable Evolutionary Data Synthesis for Verifiable Learning
• Diagnosing Representation Dynamics in NER Model Extension
• Attracting Commercial Artificial Intelligence Firms to Support National Security through Collaborative Contracts
• From Charts to Code: A Hierarchical Benchmark for Multimodal Models
• From Observations to Parameters: Detecting Changepoint in Nonlinear Dynamics with Simulation-based Inference
• AtlasKV: Augmenting LLMs with Billion-Scale Knowledge Graphs in 20GB VRAM
• XDXD: End-to-end crystal structure determination with low resolution X-ray diffraction
• UniRL-Zero: Reinforcement Learning on Unified Models with Joint Language Model and Diffusion Model Experts
• The Integration of Artificial Intelligence in Undergraduate Medical Education in Spain: Descriptive Analysis and International Perspectives
• Believe It or Not: How Deeply do LLMs Believe Implanted Facts?
• Trust in foundation models and GenAI: A geographic perspective
• Intuitionistic $j$-Do-Calculus in Topos Causal Models
• PLAGUE: Plug-and-play framework for Lifelong Adaptive Generation of Multi-turn Exploits
• Studying the Effects of Robot Intervention on School Shooters in Virtual Reality
• Universal Spectral Tokenization via Self-Supervised Panchromatic Representation Learning
• SimBA: Simplifying Benchmark Analysis Using Performance Matrices Alone
• BadScientist: Can a Research Agent Write Convincing but Unsound Papers that Fool LLM Reviewers?
• Is Multilingual LLM Watermarking Truly Multilingual? A Simple Back-Translation Solution
• DynaQuery: A Self-Adapting Framework for Querying Structured and Multimodal Data
• From Local to Global: Revisiting Structured Pruning Paradigms for Large Language Models
• SAVANT: Semantic Analysis with Vision-Augmented Anomaly deTection
• TriggerNet: A Novel Explainable AI Framework for Red Palm Mite Detection and Multi-Model Comparison and Heuristic-Guided Annotation
• Cross-Domain Long-Term Forecasting: Radiation Dose from Sparse Neutron Sensor via Spatio-Temporal Operator Network
• Language Models as Semantic Augmenters for Sequential Recommenders
• Measure-Theoretic Anti-Causal Representation Learning
• Adaptive Divergence Regularized Policy Optimization for Fine-tuning Generative Models
• SPACeR: Self-Play Anchoring with Centralized Reference Models
• Fine-tuning Flow Matching Generative Models with Intermediate Feedback
• R2L: Reliable Reinforcement Learning: Guaranteed Return & Reliable Policies in Reinforcement Learning
• Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth
• RL-Driven Security-Aware Resource Allocation Framework for UAV-Assisted O-RAN
• R2BC: Multi-Agent Imitation Learning from Single-Agent Demonstrations
• Accelerating Vision Transformers with Adaptive Patch Sizes
• Enhancing mortality prediction in cardiac arrest ICU patients through meta-modeling of structured clinical data from MIMIC-IV
• From AutoRecSys to AutoRecLab: A Call to Build, Evaluate, and Govern Autonomous Recommender-Systems Research Labs
• Latent Discrete Diffusion Models
• SafeCoop: Unravelling Full Stack Safety in Agentic Collaborative Driving
• Automatic Prompt Generation via Adaptive Selection of Prompting Techniques
• ActivationReasoning: Logical Reasoning in Latent Activation Spaces
• VelocityNet: Real-Time Crowd Anomaly Detection via Person-Specific Velocity Analysis
• RadDiagSeg-M: A Vision Language Model for Joint Diagnosis and Multi-Target Segmentation in Radiology
• Contrastive Decoding Mitigates Score Range Bias in LLM-as-a-Judge
• VLSU: Mapping the Limits of Joint Multimodal Understanding for AI Safety
• The Emergence of Complex Behavior in Large-Scale Ecological Environments
• EVER: Edge-Assisted Auto-Verification for Mobile MR-Aided Operation
• Scaling Laws Meet Model Architecture: Toward Inference-Efficient LLMs
• Finding the Sweet Spot: Optimal Data Augmentation Ratio for Imbalanced Credit Scoring Using ADASYN
• Hyperbolic Space Learning Method Leveraging Temporal Motion Priors for Human Mesh Recovery
• DelvePO: Direction-Guided Self-Evolving Framework for Flexible Prompt Optimization
• NTKMTL: Mitigating Task Imbalance in Multi-Task Learning from Neural Tangent Kernel Perspective
• Learning under Quantization for High-Dimensional Linear Regression
• SPIKE: Stable Physics-Informed Kernel Evolution Method for Solving Hyperbolic Conservation Laws
• Latent-Info and Low-Dimensional Learning for Human Mesh Recovery and Parallel Optimization
• StreamingTOM: Streaming Token Compression for Efficient Video Understanding
• Text or Pixels? It Takes Half: On the Token Efficiency of Visual Text Inputs in Multimodal LLMs
• From Retrieval to Generation: Unifying External and Parametric Knowledge for Medical Question Answering
• Higher Embedding Dimension Creates a Stronger World Model for a Simple Sorting Task
• MoMaGen: Generating Demonstrations under Soft and Hard Constraints for Multi-Step Bimanual Mobile Manipulation
• Scalable, Explainable and Provably Robust Anomaly Detection with One-Step Flow Matching
• PGTT: Phase-Guided Terrain Traversal for Perceptive Legged Locomotion
• S2AP: Score-space Sharpness Minimization for Adversarial Pruning
• MENTOR: A Reinforcement Learning Framework for Model Enhancement via Teacher-Optimized Rewards in Small Models
• Automated Wicket-Taking Delivery Segmentation and Weakness Detection in Cricket Videos Using OCR-Guided YOLOv8 and Trajectory Modeling
• Learning from N-Tuple Data with M Positive Instances: Unbiased Risk Estimation and Theoretical Guarantees
• On AI Verification in Open RAN
• Optimistic Higher-Order Superposition
• ScaleNet: Scaling up Pretrained Neural Networks with Incremental Parameters
• ImageGem: In-the-wild Generative Image Interaction Dataset for Generative Model Personalization
• DeLoad: Demand-Driven Short-Video Preloading with Scalable Watch-Time Estimation
• Simple and Efficient Heterogeneous Temporal Graph Neural Network
• CodeRL+: Improving Code Generation via Reinforcement with Execution Semantics Alignment
• Benchmarking Fairness-aware Graph Neural Networks in Knowledge Graphs
• One Size Fits All? A Modular Adaptive Sanitization Kit (MASK) for Customizable Privacy-Preserving Phone Scam Detection
• Zero-Shot Vehicle Model Recognition via Text-Based Retrieval-Augmented Generation
• Pay Attention to the Triggers: Constructing Backdoors That Survive Distillation
• EfficientNav: Towards On-Device Object-Goal Navigation with Navigation Map Caching and Retrieval
• RAISE: A Unified Framework for Responsible AI Scoring and Evaluation
• WebDevJudge: Evaluating (M)LLMs as Critiques for Web Development Quality
• Large language models for folktale type automation based on motifs: Cinderella case study
• Kaleido: Open-Sourced Multi-Subject Reference Video Generation Model
• The Cost-Benefit of Interdisciplinarity in AI for Mental Health
• A Rectification-Based Approach for Distilling Boosted Trees into Decision Trees
• Think with 3D: Geometric Imagination Grounded Spatial Reasoning from Limited Views
• C-SWAP: Explainability-Aware Structured Pruning for Efficient Neural Networks Compression
• {\epsilon}-Seg: Sparsely Supervised Semantic Segmentation of Microscopy Data
• Binary Quadratic Quantization: Beyond First-Order Quantization for Real-Valued Matrix Compression
• Reasoning Language Model Inference Serving Unveiled: An Empirical Study
• Exploring Membership Inference Vulnerabilities in Clinical Large Language Models
• Fetch.ai: An Architecture for Modern Multi-Agent Systems
• Preference-based Reinforcement Learning beyond Pairwise Comparisons: Benefits of Multiple Options
• Causally Perturbed Fairness Testing
• HarmNet: A Framework for Adaptive Multi-Turn Jailbreak Attacks on Large Language Models
• Verifiable Accuracy and Abstention Rewards in Curriculum RL to Alleviate Lost-in-Conversation
• Computational Foundations for Strategic Coopetition: Formalizing Interdependence and Complementarity
• Online SFT for LLM Reasoning: Surprising Effectiveness of Self-Tuning without Rewards
• Fine-Tuned Thoughts: Leveraging Chain-of-Thought Reasoning for Industrial Asset Health Monitoring
• An Explainable Hybrid AI Framework for Enhanced Tuberculosis and Symptom Detection
• Actor-Free Continuous Control via Structurally Maximizable Q-Functions
• Towards Faithful and Controllable Personalization via Critique-Post-Edit Reinforcement Learning
• DP$^2$O-SR: Direct Perceptual Preference Optimization for Real-World Image Super-Resolution
• Lyapunov-Aware Quantum-Inspired Reinforcement Learning for Continuous-Time Vehicle Control: A Feasibility Study
• Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Model
• LightMem: Lightweight and Efficient Memory-Augmented Generation
• How Do LLMs Use Their Depth?
• Grasp Any Region: Towards Precise, Contextual Pixel Understanding for Multimodal LLMs
• Activation Manifold Projection: Liberating Task-Specific Behaviors from LLM Architectures
• Beyond More Context: Retrieval Diversity Boosts Multi-Turn Intent Understanding
• FABRIC: Framework for Agent-Based Realistic Intelligence Creation
• OPTAGENT: Optimizing Multi-Agent LLM Interactions Through Verbal Reinforcement Learning for Enhanced Reasoning
• Subject-Event Ontology Without Global Time: Foundations and Execution Semantics
• CompactPrompt: A Unified Pipeline for Prompt Data Compression in LLM Workflows
• Planned Diffusion
• SMaRT: Select, Mix, and ReinvenT - A Strategy Fusion Framework for LLM-Driven Reasoning and Planning
• Measuring Reasoning in LLMs: a New Dialectical Angle
• Learning from Generalization Patterns: An Evaluation-Driven Approach to Enhanced Data Augmentation for Fine-Tuning Small Language Models
• Annotating the Chain-of-Thought: A Behavior-Labeled Dataset for AI Safety
• LLM-Based Multi-Agent System for Simulating and Analyzing Marketing and Consumer Behavior
• Saber: An Efficient Sampling with Adaptive Acceleration and Backtracking Enhanced Remasking for Diffusion Language Model
• AgentChangeBench: A Multi-Dimensional Evaluation Framework for Goal-Shift Robustness in Conversational AI
• Local Coherence or Global Validity? Investigating RLVR Traces in Math Domains
• FST.ai 2.0: An Explainable AI Ecosystem for Fair, Fast, and Inclusive Decision-Making in Olympic and Paralympic Taekwondo
• A Definition of AGI
• ssToken: Self-modulated and Semantic-aware Token Selection for LLM Fine-tuning
• Illusions of reflection: open-ended task reveals systematic failures in Large Language Models' reflective reasoning
• Genesis: Evolving Attack Strategies for LLM Web Agent Red-Teaming
• Earth AI: Unlocking Geospatial Insights with Foundation Models and Cross-Modal Reasoning
• ShortcutBreaker: Low-Rank Noisy Bottleneck with Global Perturbation Attention for Multi-Class Unsupervised Anomaly Detection
• Memory-Augmented State Machine Prompting: A Novel LLM Agent Framework for Real-Time Strategy Games
• Heterogeneous Adversarial Play in Interactive Environments
• Deep Learning-Based Control Optimization for Glass Bottle Forming
• Med-VRAgent: A Framework for Medical Visual Reasoning-Enhanced Agents
• Automated urban waterlogging assessment and early warning through a mixture of foundation models
• AlphaOPT: Formulating Optimization Programs with Self-Improving LLM Experience Library
• PlanU: Large Language Model Decision Making through Planning under Uncertainty
• CircuitSeer: Mining High-Quality Data by Probing Mathematical Reasoning Circuits in LLMs
• Probabilistic Modeling of Intentions in Socially Intelligent LLM Agents
• LAFA: Agentic LLM-Driven Federated Analytics over Decentralized Data Sources
• StarBench: A Turn-Based RPG Benchmark for Agentic Multimodal Decision-Making and Information Seeking
• AndroidControl-Curated: Revealing the True Potential of GUI Agents through Benchmark Purification
• Crucible: Quantifying the Potential of Control Algorithms through LLM Agents
• Counterfactual Reasoning for Steerable Pluralistic Value Alignment of Large Language Models
• Physics-guided Emulators Reveal Resilience and Fragility under Operational Latencies and Outages
• SOCIA-Nabla: Textual Gradient Meets Multi-Agent Orchestration for Automated Simulator Generation
• Extracting alignment data in open models
• QuantEvolve: Automating Quantitative Strategy Discovery through Multi-Agent Evolutionary Framework
• VAR: Visual Attention Reasoning via Structured Search and Backtracking
• Leveraging Association Rules for Better Predictions and Better Explanations
• Comparative Expressivity for Structured Argumentation Frameworks with Uncertain Rules and Premises
• Query Decomposition for RAG: Balancing Exploration-Exploitation
• Sherlock Your Queries: Learning to Ask the Right Questions for Dialogue-Based Retrieval
• Seg the HAB: Language-Guided Geospatial Algae Bloom Reasoning and Segmentation
• Decoding Funded Research: Comparative Analysis of Topic Models and Uncovering the Effect of Gender and Geographic Location
• Visual Space Optimization for Zero-shot Learning
• LLM Assisted Alpha Fairness for 6 GHz WiFi and NR_U Coexistence: An Agentic Orchestrator for Throughput, Energy, and SLA
• A Biophysical-Model-Informed Source Separation Framework For EMG Decomposition
• Carbon-Aware Orchestration of Integrated Satellite Aerial Terrestrial Networks via Digital Twin
• Speak to a Protein: An Interactive Multimodal Co-Scientist for Protein Analysis
• Multi-Agent Design Assistant for the Simulation of Inertial Fusion Energy
• Synthetic EEG Generation using Diffusion Models for Motor Imagery Tasks
• Brain-Language Model Alignment: Insights into the Platonic Hypothesis and Intermediate-Layer Advantage
• GRETEL: A Goal-driven Retrieval and Execution-based Trial Framework for LLM Tool Selection Enhancing
• Modeling Layered Consciousness with Multi-Agent Large Language Models
• MAT-Agent: Adaptive Multi-Agent Training Optimization
• CARLE: A Hybrid Deep-Shallow Learning Framework for Robust and Explainable RUL Estimation of Rolling Element Bearings
• Pre to Post-Treatment Glioblastoma MRI Prediction using a Latent Diffusion Model
• Deploying Atmospheric and Oceanic AI Models on Chinese Hardware and Framework: Migration Strategies, Performance Optimization and Analysis
• MUSE: Model-based Uncertainty-aware Similarity Estimation for zero-shot 2D Object Detection and Segmentation
• A Survey of Recursive and Recurrent Neural Networks
• Auditing and Mitigating Bias in Gender Classification Algorithms: A Data-Centric Approach
• Repairing Tool Calls Using Post-tool Execution Reflection and RAG
• 3D Weakly Supervised Semantic Segmentation via Class-Aware and Geometry-Guided Pseudo-Label Refinement
• DRL-Based Resource Allocation for Energy-Efficient IRS-Assisted UAV Spectrum Sharing Systems
• Decoding Listeners Identity: Person Identification from EEG Signals Using a Lightweight Spiking Transformer
• Outraged AI: Large language models prioritise emotion over cost in fairness enforcement
• POPI: Personalizing LLMs via Optimized Natural Language Preference Inference