AI Research News Feeds for October 6th, 2025

AI RESEARCH PAPERS & ACADEMIC SOURCES

• So-Fake: Benchmarking and Explaining Social Media Image Forgery Detection
• Revisiting Reweighted Risk for Calibration: AURC, Focal, and Inverse Focal Loss
• RichControl: Structure- and Appearance-Rich Training-Free Spatial Control for Text-to-Image Generation
• WaveNet-SF: A Hybrid Network for Retinal Disease Detection Based on Wavelet Transform in Spatial-Frequency Domain
• GCVAMD: A Modified CausalVAE Model for Causal Age-related Macular Degeneration Risk Factor Detection and Prediction
• PyRadiomics-cuda: a GPU-accelerated 3D features extraction from medical images within PyRadiomics
• Neural Posterior Estimation with Autoregressive Tiling for Detecting Objects in Astronomical Images
• Filter-Guided Diffusion for Controllable Image Generation
• Comparing YOLOv8 and Mask R-CNN for instance segmentation in complex orchard environments
• A Survey of Defenses against AI-generated Visual Media: Detection, Disruption, and Authentication
• Toward a Holistic Evaluation of Robustness in CLIP Models
• Fine-grained Abnormality Prompt Learning for Zero-shot Anomaly Detection
• Ranked from Within: Ranking Large Multimodal Models Without Labels
• SoccerSynth-Detection: A Synthetic Dataset for Soccer Player Detection
• Gate-Shift-Pose: Enhancing Action Recognition in Sports with Skeleton Information
• EFC++: Elastic Feature Consolidation with Prototype Re-balancing for Cold Start Exemplar-free Incremental Learning
• Vehicle-Scene Interaction: A Text-Driven 3D Lidar Place Recognition Method for Autonomous Driving
• Enhancing Monocular Height Estimation via Sparse LiDAR-Guided Correction
• Not every day is a sunny day: Synthetic cloud injection for deep land cover segmentation robustness evaluation across data sources
• PocketSR: The Super-Resolution Expert in Your Pocket Mobiles
• InsideOut: An EfficientNetV2-S Based Deep Learning Framework for Robust Multi-Class Facial Emotion Recognition
• Latent Diffusion Unlearning: Protecting Against Unauthorized Personalization Through Trajectory Shifted Perturbations
• GeoComplete: Geometry-Aware Diffusion for Reference-Driven Image Completion
• Taming Text-to-Sounding Video Generation via Advanced Modality Condition and Interaction
• Dynamic Prompt Generation for Interactive 3D Medical Image Segmentation Training
• Product-Quantised Image Representation for High-Quality Image Synthesis
• Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraft
• MonSTeR: a Unified Model for Motion, Scene, Text Retrieval
• MIXER: Mixed Hyperspherical Random Embedding Neural Network for Texture Recognition
• LEAML: Label-Efficient Adaptation to Out-of-Distribution Visual Tasks for Multimodal Large Language Models
• A UAV-Based VNIR Hyperspectral Benchmark Dataset for Landmine and UXO Detection
• Image Enhancement Based on Pigment Representation
• Retrv-R1: A Reasoning-Driven MLLM Framework for Universal and Efficient Multimodal Retrieval
• Bayesian Test-time Adaptation for Object Recognition and Detection with Vision-language Models
• AdaRD-key: Adaptive Relevance-Diversity Keyframe Sampling for Long-form Video understanding
• Reasoning Riddles: How Explainability Reveals Cognitive Limits in Vision-Language Models
• OTR: Synthesizing Overlay Text Dataset for Text Removal
• Med-K2N: Flexible K-to-N Modality Translation for Medical Image Synthesis
• One Patch to Caption Them All: A Unified Zero-Shot Captioning Framework
• Training-Free Out-Of-Distribution Segmentation With Foundation Models
• Don't Just Chase "Highlighted Tokens" in MLLMs: Revisiting Visual Holistic Context Retention
• Zero-Shot Robustness of Vision Language Models Via Confidence-Aware Weighting
• Flip Distribution Alignment VAE for Multi-Phase MRI Synthesis
• TIT-Score: Evaluating Long-Prompt Based Text-to-Image Alignment via Text-to-Image-to-Text Consistency
• Towards Scalable and Consistent 3D Editing
• Exploring OCR-augmented Generation for Bilingual VQA
• PhysHMR: Learning Humanoid Control Policies from Vision for Physically Plausible Human Motion Reconstruction
• Unlocking the power of partnership: How humans and machines can work together to improve face recognition
• PEO: Training-Free Aesthetic Quality Enhancement in Pre-Trained Text-to-Image Diffusion Models with Prompt Embedding Optimization
• Ego-Exo 3D Hand Tracking in the Wild with a Mobile Multi-Camera Rig
• Input-Aware Sparse Attention for Real-Time Co-Speech Video Generation
• Deep Generative Continual Learning using Functional LoRA: FunLoRA
• Sequence-Preserving Dual-FoV Defense for Traffic Sign and Light Recognition in Autonomous Vehicles
• Smart-GRPO: Smartly Sampling Noise for Efficient RL of Flow-Matching Models
• MoGIC: Boosting Motion Generation via Intention Understanding and Visual Context
• From Tokens to Nodes: Semantic-Guided Motion Control for Dynamic 3D Gaussian Splatting
• Net2Net: When Un-trained Meets Pre-trained Networks for Robust Real-World Denoising
• On the Diminishing Returns of Complex Robust RAG Training in the Era of Powerful LLMs
• BottleHumor: Self-Informed Humor Explanation using the Information Bottleneck Principle
• From TOWER to SPIRE: Adding the Speech Modality to a Text-Only LLM
• Same evaluation, more tokens: On the effect of input length for machine translation evaluation using Large Language Models
• EconWebArena: Benchmarking Autonomous Agents on Economic Tasks in Realistic Web Environments
• Same Task, Different Circuits: Disentangling Modality-Specific Mechanisms in VLMs
• Did I Faithfully Say What I Thought? Bridging the Gap Between Neural Activity and Self-Explanations in Large Language Models
• Query-Level Uncertainty in Large Language Models
• Triadic Multi-party Voice Activity Projection for Turn-taking in Spoken Dialogue Systems
• The Prompt Makes the Person(a): A Systematic Evaluation of Sociodemographic Persona Prompting for Large Language Models
• From Long Videos to Engaging Clips: A Human-Inspired Video Editing Framework with Multimodal Narrative Understanding
• Leave No TRACE: Black-box Detection of Copyrighted Dataset Usage in Large Language Models via Watermarking
• Revisiting Direct Speech-to-Text Translation with Speech LLMs: Better Scaling than CoT Prompting?
• Semantic Similarity in Radiology Reports via LLMs and NER
• Listening or Reading? Evaluating Speech Awareness in Chain-of-Thought Speech-to-Text Translation
• SurveyBench: How Well Can LLM(-Agents) Write Academic Surveys?
• Beyond the Final Layer: Intermediate Representations for Better Multilingual Calibration in Large Language Models
• EditLens: Quantifying the Extent of AI Editing in Text
• Neural Correlates of Language Models Are Specific to Human Language
• Model-Based Ranking of Source Languages for Zero-Shot Cross-Lingual Transfer
• FocusAgent: Simple Yet Effective Ways of Trimming the Large Context of Web Agents
• SpeechCT-CLIP: Distilling Text-Image Knowledge to Speech for Voice-Native Multimodal CT Analysis
• When Names Disappear: Revealing What LLMs Actually Understand About Code
• Did Translation Models Get More Robust Without Anyone Even Noticing?
• ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference
• KnowledgeSmith: Uncovering Knowledge Updating in LLMs with Model Editing and Unlearning
• Retrieval and Augmentation of Domain Knowledge for Text-to-SQL Semantic Parsing
• Transcribe, Translate, or Transliterate: An Investigation of Intermediate Representations in Spoken Language Models
• Evaluation Framework for Highlight Explanations of Context Utilisation in Language Models
• Mind the Gap: Linguistic Divergence and Adaptation Strategies in Human-LLM Assistant vs. Human-Human Interactions
• SoT: Structured-of-Thought Prompting Guides Multilingual Reasoning in Large Language Models
• Self-Improvement in Multimodal Large Language Models: A Survey
• PGMEL: Policy Gradient-based Generative Adversarial Network for Multimodal Entity Linking
• IndiCASA: A Dataset and Bias Evaluation Framework in LLMs Using Contrastive Embedding Similarity in the Indian Context
• The Path of Self-Evolving Large Language Models: Achieving Data-Efficient Learning via Intrinsic Feedback
• XTRA: Cross-Lingual Topic Modeling with Topic and Representation Alignments
• Self-Reflective Generation at Test Time
• Finding Diamonds in Conversation Haystacks: A Benchmark for Conversational Data Retrieval
• Learning to Route: A Rule-Driven Agent Framework for Hybrid-Source Retrieval-Augmented Generation
• Understanding How CodeLLMs (Mis)Predict Types with Activation Steering
• Programming with Pixels: Can Computer-Use Agents do Software Engineering?
• DeepGDel: Deep Learning-based Gene Deletion Prediction Framework for Growth-Coupled Production in Genome-Scale Metabolic Models
• On the $O(\frac{\sqrt{d}}{K^{1/4}})$ Convergence Rate of AdamW Measured by $\ell_1$ Norm
• Graph-Reward-SQL: Execution-Free Reinforcement Learning for Text-to-SQL via Graph Matching and Stepwise Reward
• Understanding the Performance Gap in Preference Learning: A Dichotomy of RLHF and DPO
• QiMeng-CodeV-R1: Reasoning-Enhanced Verilog Generation
• Risk-Sensitive Agent Compositions
• Generative Modeling of Weights: Generalization or Memorization?
• Modern Methods in Associative Memory
• FR-LUX: Friction-Aware, Regime-Conditioned Policy Optimization for Implementable Portfolio Management
• The Computational Complexity of Almost Stable Clustering with Penalties
• ReeMark: Reeb Graphs for Simulating Patterns of Life in Spatiotemporal Trajectories
• Improving Online-to-Nonconvex Conversion for Smooth Optimization via Double Optimism
• Automatic Generation of Digital Twins for Network Testing
• Joint Bidding on Intraday and Frequency Containment Reserve Markets
• Cache-to-Cache: Direct Semantic Communication Between Large Language Models
• The Challenges of Hyperparameter Tuning for Accurate Causal Effect Estimation
• Wasserstein Bounds for generative diffusion models with Gaussian tail targets
• ColNet: Collaborative Optimization in Decentralized Federated Multi-task Learning Systems
• Best Policy Learning from Trajectory Preference Feedback
• On the Effect of Sampling Diversity in Scaling LLM Inference
• To Backtrack or Not to Backtrack: When Sequential Search Limits Model Reasoning
• Heterogeneous Graph Representation of Stiffened Panels with Non-Uniform Boundary Conditions and Loads
• Unraveling Syntax: How Language Models Learn Context-Free Grammars
• Self-supervised diffusion model fine-tuning for costate initialization using Markov chain Monte Carlo
• Even Faster Kernel Matrix Linear Algebra via Density Estimation
• FLOWR.root: A flow matching based foundation model for joint multi-purpose structure-aware 3D ligand generation and affinity prediction
• Uncertainty as Feature Gaps: Epistemic Uncertainty Quantification of LLMs in Contextual Question-Answering
• Quantitative Convergence Analysis of Projected Stochastic Gradient Descent for Non-Convex Losses via the Goldstein Subdifferential
• The land use-climate change-biodiversity nexus in European islands stakeholders
• ELMF4EggQ: Ensemble Learning with Multimodal Feature Fusion for Non-Destructive Egg Quality Assessment
• SALSA-V: Shortcut-Augmented Long-form Synchronized Audio from Videos
• Mechanistic Interpretability of Code Correctness in LLMs via Sparse Autoencoders
• Scalable Quantum Optimisation using HADOF: Hamiltonian Auto-Decomposition Optimisation Framework
• oRANS: Online optimisation of RANS machine learning models with embedded DNS data generation
• Q-Learning with Shift-Aware Upper Confidence Bound in Non-Stationary Reinforcement Learning
• PRISM-Physics: Causal DAG-Based Process Evaluation for Physics Reasoning
• Superposition disentanglement of neural representations reveals hidden alignment
• Estimation of Resistance Training RPE using Inertial Sensors and Electromyography
• To Distill or Decide? Understanding the Algorithmic Trade-off in Partially Observable Reinforcement Learning
• Low-probability Tokens Sustain Exploration in Reinforcement Learning with Verifiable Reward
• WEE-Therapy: A Mixture of Weak Encoders Framework for Psychological Counseling Dialogue Analysis
• Can Prompts Rewind Time for LLMs? Evaluating the Effectiveness of Prompted Knowledge Cutoffs
• An Senegalese Legal Texts Structuration Using LLM-augmented Knowledge Graph
• Modeling the language cortex with form-independent and enriched representations of sentence meaning reveals remarkable semantic abstractness
• An Encoder-Decoder Network for Beamforming over Sparse Large-Scale MIMO Channels
• Uncertainty-Aware Answer Selection for Improved Reasoning in Multi-LLM Systems
• The Equilibrium Response of Atmospheric Machine-Learning Models to Uniform Sea Surface Temperature Warming
• Words That Make Language Models Perceive
• Confidence and Dispersity as Signals: Unsupervised Model Evaluation and Ranking
• Distributional Inverse Reinforcement Learning
• Lightweight Transformer for EEG Classification via Balanced Signed Graph Algorithm Unrolling
• Bayesian E(3)-Equivariant Interatomic Potential with Iterative Restratification of Many-body Message Passing
• Bootstrap Learning for Combinatorial Graph Alignment with Sequential GNNs
• Adaptive Node Feature Selection For Graph Neural Networks
• AdaBet: Gradient-free Layer Selection for Efficient Training of Deep Neural Networks
• Real Time Headway Predictions in Urban Rail Systems and Implications for Service Control: A Deep Learning Approach
• Enhancing XAI Narratives through Multi-Narrative Refinement and Knowledge Distillation
• Taming Imperfect Process Verifiers: A Sampling Perspective on Backtracking
• Mixture of Many Zero-Compute Experts: A High-Rate Quantization Theory Perspective
• Calibrated Uncertainty Sampling for Active Learning
• FTTE: Federated Learning on Resource-Constrained Devices
• TokenFlow: Responsive LLM Text Streaming Serving under Request Burst via Preemptive Scheduling
• Curl Descent: Non-Gradient Learning Dynamics with Sign-Diverse Plasticity
• A Granular Study of Safety Pretraining under Model Abliteration
• Optimal Rates for Generalization of Gradient Descent for Deep ReLU Classification
• Mitigating Spurious Correlation via Distributionally Robust Learning with Hierarchical Ambiguity Sets
• Online Learning in the Random Order Model
• FlexiQ: Adaptive Mixed-Precision Quantization for Latency/Accuracy Trade-Offs in Deep Neural Networks
• The Curious Case of In-Training Compression of State Space Models
• Multi-scale Autoregressive Models are Laplacian, Discrete, and Latent Diffusion Models in Disguise
• Subject-Adaptive Sparse Linear Models for Interpretable Personalized Health Prediction from Multimodal Lifelog Data
• RoiRL: Efficient, Self-Supervised Reasoning with Offline Iterative Reinforcement Learning
• Learning Explicit Single-Cell Dynamics Using ODE Representations
• RAxSS: Retrieval-Augmented Sparse Sampling for Explainable Variable-Length Medical Time Series Classification
• ContextFlow: Context-Aware Flow Matching For Trajectory Inference From Spatial Omics Data
• AttentiveGRUAE: An Attention-Based GRU Autoencoder for Temporal Clustering and Behavioral Characterization of Depression from Wearable Data
• On The Expressive Power of GNN Derivatives
• Geospatial Machine Learning Libraries
• Use the Online Network If You Can: Towards Fast and Stable Reinforcement Learning
• Towards CONUS-Wide ML-Augmented Conceptually-Interpretable Modeling of Catchment-Scale Precipitation-Storage-Runoff Dynamics
• TabImpute: Accurate and Fast Zero-Shot Missing-Data Imputation with a Pre-Trained Transformer
• HyperAdaLoRA: Accelerating LoRA Rank Allocation During Training via Hypernetworks without Sacrificing Performance
• Optimal Characteristics of Inspection Vehicle for Drive-by Bridge Inspection
• Topological Invariance and Breakdown in Learning
• EvoSpeak: Large Language Models for Interpretable Genetic Programming-Evolved Heuristics
• Accuracy Law for the Future of Deep Time Series Forecasting
• Dale meets Langevin: A Multiplicative Denoising Diffusion Model
• Hybrid-Collaborative Augmentation and Contrastive Sample Adaptive-Differential Awareness for Robust Attributed Graph Clustering
• OpenTSLM: Time-Series Language Models for Reasoning over Multivariate Medical Text- and Time-Series Data
• Assessing the Potential for Catastrophic Failure in Dynamic Post-Training Quantization
• SAGE: Streaming Agreement-Driven Gradient Sketches for Representative Subset Selection
• Uncertainty-Guided Model Selection for Tabular Foundation Models in Biomolecule Efficacy Prediction
• Improved Robustness of Deep Reinforcement Learning for Control of Time-Varying Systems by Bounded Extremum Seeking
• Beyond Imitation: Recovering Dense Rewards from Demonstrations
• In-memory Training on Analog Devices with Limited Conductance States via Multi-tile Residual Learning
• Graph Generation with Spectral Geodesic Flow Matching
• Model-brain comparison using inter-animal transforms
• STORI: A Benchmark and Taxonomy for Stochastic Environments
• PropRAG: Guiding Retrieval with Beam Search over Proposition Paths
• AlignDiT: Multimodal Aligned Diffusion Transformer for Synchronized Speech Generation
• Continuous Thought Machines
• A Survey of Deep Learning for Complex Speech Spectrograms
• OT Score: An OT based Confidence Score for Source Free Unsupervised Domain Adaptation
• Pre-training Limited Memory Language Models with Internal and External Knowledge
• NeSyGeo: A Neuro-Symbolic Framework for Multimodal Geometric Reasoning Data Generation
• Manipulating 3D Molecules in a Fixed-Dimensional E(3)-Equivariant Latent Space
• SP-VLA: A Joint Model Scheduling and Token Pruning Approach for VLA Model Acceleration
• A Survey of Pun Generation: Datasets, Evaluations and Methodologies
• Unified Domain Adaptive Semantic Segmentation
• RACCooN: A Versatile Instructional Video Editing Framework with Auto-Generated Narratives
• Optimizing Container Loading and Unloading through Dual-Cycling and Dockyard Rehandle Reduction Using a Hybrid Genetic Algorithm
• Inferring Pluggable Types with Machine Learning
• CBVLM: Training-free Explainable Concept-based Large Vision Language Models for Medical Image Classification
• MarketSenseAI 2.0: Enhancing Stock Analysis through LLM Agents
• Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs
• L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning
• DatawiseAgent: A Notebook-Centric LLM Agent Framework for Adaptive and Robust Data Science Automation
• Towards Quantifying Long-Range Interactions in Graph Machine Learning: a Large Graph Dataset and a Measurement
• Verbosity Tradeoffs and the Impact of Scale on the Faithfulness of LLM Self-Explanations
• Not a nuisance but a useful heuristic: Outlier dimensions favor frequent tokens in language models
• Activated LoRA: Fine-tuned LLMs for Intrinsics
• Abstain and Validate: A Dual-LLM Policy for Reducing Noise in Agentic Program Repair
• Self-Anchor: Large Language Model Reasoning via Step-by-step Attention Alignment
• Test-Time Defense Against Adversarial Attacks via Stochastic Resonance of Latent Ensembles
• Improving GUI Grounding with Explicit Position-to-Coordinate Mapping
• Reward Models are Metrics in a Trench Coat
• Improved Monte Carlo Planning via Causal Disentanglement for Structurally-Decomposed Markov Decision Processes
• MIRROR: Modular Internal Processing for Personalized Safety in LLM Dialogue
• V2X-UniPool: Unifying Multimodal Perception and Knowledge Reasoning for Autonomous Driving
• LayerCake: Token-Aware Contrastive Decoding within Large Language Model Layers
• THOR: Tool-Integrated Hierarchical Optimization via RL for Mathematical Reasoning
• ZeroShotOpt: Towards Zero-Shot Pretrained Models for Efficient Black-Box Optimization
• Semantic Differentiation in Speech Emotion Recognition: Insights from Descriptive and Expressive Speech Roles
• Comparative Analysis of Parameterized Action Actor-Critic Reinforcement Learning Algorithms for Web Search Match Plan Generation
• A Unified Deep Reinforcement Learning Approach for Close Enough Traveling Salesman Problem
• A Study of Neural Polar Decoders for Communication
• What Drives Compositional Generalization in Visual Generative Models?
• HAVIR: HierArchical Vision to Image Reconstruction using CLIP-Guided Versatile Diffusion
• Signature-Informed Transformer for Asset Allocation
• Stimulus-Voltage-Based Prediction of Action Potential Onset Timing: Classical vs. Quantum-Inspired Approaches
• SpineBench: A Clinically Salient, Level-Aware Benchmark Powered by the SpineMed-450k Corpus
• UniShield: An Adaptive Multi-Agent Framework for Unified Forgery Image Detection and Localization
• Topic Modeling as Long-Form Generation: Can Long-Context LLMs revolutionize NTM via Zero-Shot Prompting?
• Wave-GMS: Lightweight Multi-Scale Generative Model for Medical Image Segmentation
• Constraint Satisfaction Approaches to Wordle: Novel Heuristics and Cross-Lexicon Validation
• Representing Beauty: Towards a Participatory but Objective Latent Aesthetics
• Global Convergence of Policy Gradient for Entropy Regularized Linear-Quadratic Control with multiplicative noise
• FinReflectKG - MultiHop: Financial QA Benchmark for Reasoning with Knowledge Graph Evidence
• FeDABoost: Fairness Aware Federated Learning with Adaptive Boosting
• Multimodal Carotid Risk Stratification with Large Vision-Language Models: Benchmarking, Fine-Tuning, and Clinical Insights
• Ergodic Risk Measures: Towards a Risk-Aware Foundation for Continual Reinforcement Learning
• Corrosion Risk Estimation for Heritage Preservation: An Internet of Things and Machine Learning Approach Using Temperature and Humidity
• From high-frequency sensors to noon reports: Using transfer learning for shaft power prediction in maritime
• BrainIB++: Leveraging Graph Neural Networks and Information Bottleneck for Functional Brain Biomarkers in Schizophrenia
• Learning Robust Diffusion Models from Imprecise Supervision
• Investigating The Smells of LLM Generated Code
• When and Where do Events Switch in Multi-Event Video Generation?
• TravelBench : Exploring LLM Performance in Low-Resource Domains
• SAE-RNA: A Sparse Autoencoder Model for Interpreting RNA Language Model Representations
• Hierarchical Generalized Category Discovery for Brain Tumor Classification in Digital Pathology
• Fusing Multi- and Hyperspectral Satellite Data for Harmful Algal Bloom Monitoring with Self-Supervised and Hierarchical Deep Learning
• Align Your Query: Representation Alignment for Multimodality Medical Object Detection
• MaskCD: Mitigating LVLM Hallucinations by Image Head Masked Contrastive Decoding
• Pareto-optimal Non-uniform Language Generation
• OptunaHub: A Platform for Black-Box Optimization
• Relevance-Aware Thresholding in Online Conformal Prediction for Time Series
• Dissecting Transformers: A CLEAR Perspective towards Green AI
• A Computational Framework for Interpretable Text-Based Personality Assessment from Social Media
• Evaluating Large Language Models for IUCN Red List Species Information
• Knowledge-Aware Modeling with Frequency Adaptive Learning for Battery Health Prognostics
• Flamed-TTS: Flow Matching Attention-Free Models for Efficient Generating and Dynamic Pacing Zero-shot Text-to-Speech
• Knowledge-Graph Based RAG System Evaluation Framework
• Oracle-RLAIF: An Improved Fine-Tuning Framework for Multi-modal Video Models through Reinforcement Learning from Ranking Feedback
• How Confident are Video Models? Empowering Video Models to Express their Uncertainty
• MINERVA: Mutual Information Neural Estimation for Supervised Feature Selection
• Automatic Building Code Review: A Case Study
• TutorBench: A Benchmark To Assess Tutoring Capabilities Of Large Language Models
• HALO: Memory-Centric Heterogeneous Accelerator with 2.5D Integration for Low-Batch LLM Inference
• To Compress or Not? Pushing the Frontier of Lossless GenAI Model Weights Compression with Exponent Concentration
• Can Data-Driven Dynamics Reveal Hidden Physics? There Is A Need for Interpretable Neural Operators
• Fine-Tuning Diffusion Models via Intermediate Distribution Shaping
• RAMAC: Multimodal Risk-Aware Offline Reinforcement Learning and the Role of Behavior Regularization
• Time-To-Inconsistency: A Survival Analysis of Large Language Model Robustness to Adversarial Attacks
• Fully automated inverse co-optimization of templates and block copolymer blending recipes for DSA lithography
• CWM: An Open-Weights LLM for Research on Code Generation with World Models
• Linear RNNs for autoregressive generation of long music samples
• Glaucoma Detection and Structured OCT Report Generation via a Fine-tuned Multimodal Large Language Model
• Extreme value forecasting using relevance-based data augmentation with deep learning models
• RainSeer: Fine-Grained Rainfall Reconstruction via Physics-Guided Modeling
• Cross-Platform DNA Methylation Classifier for the Eight Molecular Subtypes of Group 3 & 4 Medulloblastoma
• NEURODNAAI: Neural pipeline approaches for the advancing dna-based information storage as a sustainable digital medium using deep learning framework
• How to Train Your Advisor: Steering Black-Box LLMs with Advisor Models
• Market-Based Data Subset Selection -- Principled Aggregation of Multi-Criteria Example Utility
• CLARITY: Clinical Assistant for Routing, Inference, and Triage
• Litespark Technical Report: High-Throughput, Energy-Efficient LLM Training Framework
• From Pixels to Factors: Learning Independently Controllable State Variables for Reinforcement Learning
• PHORECAST: Enabling AI Understanding of Public Health Outreach Across Populations
• Breaking the MoE LLM Trilemma: Dynamic Expert Clustering with Structured Compression
• Small Language Models for Curriculum-based Guidance
• mini-vec2vec: Scaling Universal Geometry Alignment with Linear Transformations
• LLMSQL: Upgrading WikiSQL for the LLM Era of Text-to-SQL
• Language, Culture, and Ideology: Personalizing Offensiveness Detection in Political Tweets with Reasoning LLMs
• Evaluating Bias in Spoken Dialogue LLMs for Real-World Decisions and Recommendations
• DiffuSpec: Unlocking Diffusion Language Models for Speculative Decoding
• Emission-GPT: A domain-specific language model agent for knowledge retrieval, emission inventory and data analysis
• Spiral of Silence in Large Language Model Agents
• ChunkLLM: A Lightweight Pluggable Framework for Accelerating LLMs Inference
• A Cross-Lingual Analysis of Bias in Large Language Models Using Romanian History
• Beyond Manuals and Tasks: Instance-Level Context Learning for LLM Agents
• Training Dynamics of Parametric and In-Context Knowledge Utilization in Language Models
• Pretraining with hierarchical memories: separating long-tail and common knowledge
• KAME: Tandem Architecture for Enhancing Knowledge in Real-Time Speech-to-Speech Conversational AI
• AMANDA: Agentic Medical Knowledge Augmentation for Data-Efficient Medical Visual Question Answering
• SelfJudge: Faster Speculative Decoding via Self-Supervised Judge Verification
• EntropyLong: Effective Long-Context Training via Predictive Uncertainty
• Synthetic Dialogue Generation for Interactive Conversational Elicitation & Recommendation (ICER)
• Human Mobility Datasets Enriched With Contextual and Social Dimensions
• Where Did It Go Wrong? Attributing Undesirable LLM Behaviors via Representation Gradient Tracing
• FormalML: A Benchmark for Evaluating Formal Subgoal Completion in Machine Learning Theory
• KurdSTS: The Kurdish Semantic Textual Similarity
• CRACQ: A Multi-Dimensional Approach To Automated Document Assessment
• Optimizing Long-Form Clinical Text Generation with Claim-Based Rewards
• Evaluating Uncertainty Quantification Methods in Argumentative Large Language Models
• DRIFT: Learning from Abundant User Dissatisfaction in Real-World Preference Learning
• $\texttt{BluePrint}$: A Social Media User Dataset for LLM Persona Evaluation and Training
• Onto-Epistemological Analysis of AI Explanations
• From Facts to Foils: Designing and Evaluating Counterfactual Explanations for Smart Environments
• A Study of Rule Omission in Raven's Progressive Matrices
• CoDA: Agentic Systems for Collaborative Data Visualization
• Coevolutionary Continuous Discrete Diffusion: Make Your Diffusion Language Model a Latent Reasoner
• Representation Learning for Compressed Video Action Recognition via Attentive Cross-modal Interaction with Motion Enhancement
• Hallucination reduction with CASAL: Contrastive Activation Steering For Amortized Learning
• Hallucination-Resistant, Domain-Specific Research Assistant with Self-Evaluation and Vector-Grounded Retrieval
• NCV: A Node-Wise Consistency Verification Approach for Low-Cost Structured Error Localization in LLM Reasoning
• Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents
• Take Goodhart Seriously: Principled Limit on General-Purpose AI Optimization
• Reward Model Routing in Alignment
• Consolidating Reinforcement Learning for Multimodal Discrete Diffusion Models
• BrowserArena: Evaluating LLM Agents on Real-World Web Navigation Tasks
• RefineShot: Rethinking Cinematography Understanding with Foundational Skill Evaluation
• Safe and Efficient In-Context Learning via Risk Control
• Multimodal Function Vectors for Spatial Relations
• Orchestrating Human-AI Teams: The Manager Agent as a Unifying Research Challenge
• Agentic Additive Manufacturing Alloy Discovery
• A Benchmark Study of Deep Reinforcement Learning Algorithms for the Container Stowage Planning Problem
• Multimodal Large Language Model Framework for Safe and Interpretable Grid-Integrated EVs
• Mitigating Modal Imbalance in Multimodal Reasoning
• On the Role of Temperature Sampling in Test-Time Scaling
• A Concept of Possibility for Real-World Events
• ARMs: Adaptive Red-Teaming Agent against Multimodal Models with Plug-and-Play Attacks
• Automated Constraint Specification for Job Scheduling by Regulating Generative Model with Domain-Specific Representation
• Efficient Preimage Approximation for Neural Network Certification
• Smart Contract Intent Detection with Pre-trained Programming Language Model
• XBreaking: Explainable Artificial Intelligence for Jailbreaking LLMs
• JALMBench: Benchmarking Jailbreak Vulnerabilities in Audio Language Models
• Permissioned LLMs: Enforcing Access Control in Large Language Models
• Semantic Preprocessing for LLM-based Malware Analysis
• Diffusion-aided Task-oriented Semantic Communications with Model Inversion Attack
• GATEBLEED: Exploiting On-Core Accelerator Power Gating for High Performance & Stealthy Attacks on AI
• ShikkhaChain: A Blockchain-Powered Academic Credential Verification System for Bangladesh
• Secure and Scalable Blockchain Voting: A Comparative Framework and the Role of Large Language Models
• Scam2Prompt: A Scalable Framework for Auditing Malicious Scam Endpoints in Production LLMs
• Mutual Information Guided Backdoor Mitigation for Pre-trained Encoders
• OML: A Primitive for Reconciling Open Access with Owner Control in AI Model Distribution
• Rethinking the Vulnerability of Concept Erasure and a New Method
• FinP: Fairness-in-Privacy in Federated Learning by Addressing Disparities in Privacy Risk
• Interplay between Security, Privacy and Trust in 6G-enabled Intelligent Transportation Systems
• A Bilevel Optimization Framework for Adversarial Control of Gas Pipeline Operations
• A Novel Unified Lightweight Temporal-Spatial Transformer Approach for Intrusion Detection in Drone Networks
• CST-AFNet: A dual attention-based deep learning framework for intrusion detection in IoT networks
• Automated Repair of OpenID Connect Programs (Extended Version)
• DMark: Order-Agnostic Watermarking for Diffusion Large Language Models
• WavInWav: Time-domain Speech Hiding via Invertible Neural Network
• Cheat-Penalised Quantum Weak Coin-Flipping
• LLAMAFUZZ: Large Language Model Enhanced Greybox Fuzzing
• RSFuzz: A Robustness-Guided Swarm Fuzzing Framework Based on Behavioral Constraints
• PrisonBreak: Jailbreaking Large Language Models with at Most Twenty-Five Targeted Bit-flips
• Primus: A Pioneering Collection of Open-Source Datasets for Cybersecurity LLM Training
• Activation Functions Considered Harmful: Recovering Neural Network Weights through Controlled Channels
• Using Preformed Resistive Random Access Memory to Create a Strong Physically Unclonable Function
• MALF: A Multi-Agent LLM Framework for Intelligent Fuzzing of Industrial Control Protocols
• A Statistical Method for Attack-Agnostic Adversarial Attack Detection with Compressive Sensing Comparison
• Attack via Overfitting: 10-shot Benign Fine-tuning to Jailbreak LLMs
• Improved Search-to-Decision Reduction for Random Local Functions
• SoK: Preconfirmations
• SoK: Kicking CAN Down the Road. Systematizing CAN Security Knowledge
• External Data Extraction Attacks against Retrieval-Augmented Large Language Models
• Untargeted Jailbreak Attack
• Protecting Persona Biometric Data: The Case of Facial Privacy
• TPM-Based Continuous Remote Attestation and Integrity Verification for 5G VNFs on Kubernetes
• A High-Capacity and Secure Disambiguation Algorithm for Neural Linguistic Steganography
• From Trace to Line: LLM Agent for Real-World OSS Vulnerability Localization
• Apply Bayes Theorem to Optimize IVR Authentication Process
• Hybrid Schemes of NIST Post-Quantum Cryptography Standard Algorithms and Quantum Key Distribution for Key Exchange and Digital Signature
• Selmer-Inspired Elliptic Curve Generation
• Secure and Robust Watermarking for AI-generated Images: A Comprehensive Survey
• On The Fragility of Benchmark Contamination Detection in Reasoning Models
• LLM-Generated Samples for Android Malware Detection
• PolyLink: A Blockchain Based Decentralized Edge AI Platform for LLM Inference
• Dynamic Target Attack
• Adaptive Deception Framework with Behavioral Analysis for Enhanced Cybersecurity Defense
• Rigorous Evaluation of Microarchitectural Side-Channels with Statistical Model Checking
• TLoRa: Implementing TLS Over LoRa for Secure HTTP Communication in IoT
• ToolTweak: An Attack on Tool Selection in LLM-based Agents
• Hybrid Horizons: Policy for Post-Quantum Security
• Modeling the Attack: Detecting AI-Generated Text by Quantifying Adversarial Perturbations
• Agentic-AI Healthcare: Multilingual, Privacy-First Framework with MCP Agents
• CATMark: A Context-Aware Thresholding Framework for Robust Cross-Task Watermarking in Large Language Models
• An Investigation into the Performance of Non-Contrastive Self-Supervised Learning Methods for Network Intrusion Detection
• Measuring Physical-World Privacy Awareness of Large Language Models: An Evaluation Benchmark
• Privacy in the Age of AI: A Taxonomy of Data Risks
• Bootstrapping as a Morphism: An Arithmetic Geometry Approach to Asymptotically Faster Homomorphic Encryption
• Federated Spatiotemporal Graph Learning for Passive Attack Detection in Smart Grids
• A-MemGuard: A Proactive Defense Framework for LLM-Based Agent Memory
• A Hybrid CAPTCHA Combining Generative AI with Keystroke Dynamics for Enhanced Bot Detection
• Scaling Homomorphic Applications in Deployment
• Less LLM, More Documents: Searching for Improved RAG
• AgenticRAG: Tool-Augmented Foundation Models for Zero-Shot Explainable Recommender Systems
• OpenZL: A Graph-Based Model for Compression
• Hierarchical Semantic Retrieval with Cobweb
• Geolog-IA: Conversational System for Academic Theses
• StepChain GraphRAG: Reasoning Over Knowledge Graphs for Multi-Hop Question Answering
• Grounding Large Language Models in Clinical Evidence: A Retrieval-Augmented Generation System for Querying UK NICE Clinical Guidelines
• CHORD: Customizing Hybrid-precision On-device Model for Sequential Recommendation with Device-cloud Collaboration
• cAST: Enhancing Code Retrieval-Augmented Generation with Structural Chunking via Abstract Syntax Tree
• MHier-RAG: Multi-Modal RAG for Visual-Rich Document Question-Answering via Hierarchical and Multi-Granularity Reasoning
• When Large Language Models are Reliable for Judging Empathic Communication
• Hyperparameters are all you need: Using five-step inference for an original diffusion model to generate images comparable to the latest distillation model
• Visualizing Spatial Point Clouds: A Task-Oriented Taxonomy
• GS-Share: Enabling High-fidelity Map Sharing with Incremental Gaussian Splatting
• PCG-Informed Neural Solvers for High-Resolution Homogenization of Periodic Microstructures
• FSFSplatter: Build Surface and Novel Views with Sparse-Views within 3min
• ROGR: Relightable 3D Objects using Generative Relighting
• Topological Autoencoders++: Fast and Accurate Cycle-Aware Dimensionality Reduction
• Revisiting Query Variants: The Advantage of Retrieval Over Generation of Query Variants for Effective QPP
• A Simple but Effective Elaborative Query Reformulation Approach for Natural Language Recommendation
• Open WebUI: An Open, Extensible, and Usable Interface for AI Interaction
• When Researchers Say Mental Model/Theory of Mind of AI, What Are They Really Talking About?
• "It Felt Real" Victim Perspectives on Platform Design and Longer-Running Scams
• Prototyping Digital Social Spaces through Metaphor-Driven Design: Translating Spatial Concepts into an Interactive Social Simulation
• Fostering Collective Discourse: A Distributed Role-Based Approach to Online News Commenting
• PromptMap: Supporting Exploratory Text-to-Image Generation
• VR as a "Drop-In" Well-being Tool for Knowledge Workers
• Who's Wearing? Ear Canal Biometric Key Extraction for User Authentication on Wireless Earbuds
• AutoMaAS: Self-Evolving Multi-Agent Architecture Search for Large Language Models
• AI Generated Child Sexual Abuse Material - What's the Harm?
• Reading.help: Supporting EFL Readers with Proactive and On-Demand Explanation of English Grammar and Semantics
• Vibe coding: programming through conversation with artificial intelligence
• Comparing Exploration-Exploitation Strategies of LLMs and Humans: Insights from Standard Multi-armed Bandit Experiments
• Learning High-Fidelity Robot Self-Model with Articulated 3D Gaussian Splatting
• A Benchmarking Study of Vision-based Robotic Grasping Algorithms
• Learned IMU Bias Prediction for Invariant Visual Inertial Odometry
• Latent Action Diffusion for Cross-Embodiment Manipulation
• Beyond Anthropomorphism: Enhancing Grasping and Eliminating a Degree of Freedom by Fusing the Abduction of Digits Four and Five
• An Intention-driven Lane Change Framework Considering Heterogeneous Dynamic Cooperation in Mixed-traffic Environment
• A Multi-Fidelity Control Variate Approach for Policy Gradient Estimation
• Optimal Modified Feedback Strategies in LQ Games under Control Imperfections
• HopaDIFF: Holistic-Partial Aware Fourier Conditioned Diffusion for Referring Human Action Segmentation in Multi-Person Scenarios
• Decoupling Geometry from Optimization in 2D Irregular Cutting and Packing Problems: an Open-Source Collision Detection Engine
• Vector Autoregression (VAR) of Longitudinal Sleep and Self-report Mood Data
• Learning Stability Certificate for Robotics in Real-World Environments
• MM-Nav: Multi-View VLA Model for Robust Visual Navigation via Multi-Expert Learning
• Optimal Smooth Coverage Trajectory Planning for Quadrotors in Cluttered Environment
• Simulation to Rules: A Dual-VLM Framework for Formal Visual Planning
• Conceptualizing and Modeling Communication-Based Cyberattacks on Automated Vehicles
• Periodic Event-Triggered Prescribed Time Control of Euler-Lagrange Systems under State and Input Constraints
• VERNIER: an open-source software pushing marker pose estimation down to the micrometer and nanometer scales
• A Dimension-Decomposed Learning Framework for Online Disturbance Identification in Quadrotor SE(3) Control
• Geometry Meets Vision: Revisiting Pretrained Semantics in Distilled Fields
• Mask2IV: Interaction-Centric Video Generation via Mask Trajectories
• Improving Cooperation in Collaborative Embodied AI
• S-Graphs 2.0 -- A Hierarchical-Semantic Optimization and Loop Closure for SLAM
• Active Alignments of Lens Systems with Reinforcement Learning
• Assist-as-needed Control for FES in Foot Drop Management
• Action Deviation-Aware Inference for Low-Latency Wireless Robots
• Novel UWB Synthetic Aperture Radar Imaging for Mobile Robot Mapping
• Point Cloud-Based Control Barrier Functions for Model Predictive Control in Safety-Critical Navigation of Autonomous Mobile Robots
• Metrics vs Surveys: Can Quantitative Measures Replace Human Surveys in Social Robot Navigation? A Correlation Analysis
• Single-Rod Brachiation Robot: Mechatronic Control Design and Validation of Prejump Phases
• YawSitter: Modeling and Controlling a Tail-Sitter UAV with Enhanced Yaw Control
• AI-Enhanced Kinematic Modeling of Flexible Manipulators Using Multi-IMU Sensor Fusion
• Real-Time Nonlinear Model Predictive Control of Heavy-Duty Skid-Steered Mobile Platform for Trajectory Tracking Tasks
• 3D-CovDiffusion: 3D-Aware Diffusion Policy for Coverage Path Planning
• HumanoidExo: Scalable Whole-Body Humanoid Manipulation via Wearable Exoskeleton
• Long-Term Human Motion Prediction Using Spatio-Temporal Maps of Dynamics
• Embracing Evolution: A Call for Body-Control Co-Design in Embodied Humanoid Robot
• Whisker-based Tactile Flight for Tiny Drones
• A Recipe for Efficient Sim-to-Real Transfer in Manipulation with Online Imitation-Pretrained World Models
• Efficient Optimal Path Planning in Dynamic Environments Using Koopman MPC
• SubSense: VR-Haptic and Motor Feedback for Immersive Control in Subsea Telerobotics
• UMI-on-Air: Embodiment-Aware Guidance for Embodiment-Agnostic Visuomotor Policies
• RSV-SLAM: Toward Real-Time Semantic Visual SLAM in Indoor Dynamic Environments
• Reachable Predictive Control: A Novel Control Algorithm for Nonlinear Systems with Unknown Dynamics and its Practical Applications
• Multi-robot Rigid Formation Navigation via Synchronous Motion and Discrete-time Communication-Control Optimization
• A Trajectory Generator for High-Density Traffic and Diverse Agent-Interaction Scenarios
• A $1000\times$ Faster LLM-enhanced Algorithm For Path Planning in Large-scale Grid Maps
• Team Xiaomi EV-AD VLA: Caption-Guided Retrieval System for Cross-Modal Drone Navigation - Technical Report for IROS 2025 RoboSense Challenge Track 4
• Flow with the Force Field: Learning 3D Compliant Flow Matching Policies from Force and Demonstration-Guided Simulation Data
• Work Zones challenge VLM Trajectory Planning: Toward Mitigation and Robust Autonomous Driving
• Graph Neural Networks for Transmission Grid Topology Control: Busbar Information Asymmetry and Heterogeneous Representations
• Learning Counterfactual Outcomes Under Rank Preservation
• Dynamical local Fr\'echet curve regression in manifolds
• Theoretical Investigation on Inductive Bias of Isolation Forest
• A Malliavin-Gamma calculus approach to Score Based Diffusion Generative models for random fields
• Online Decision-Focused Learning
• Restricted Spectral Gap Decomposition for Simulated Tempering Targeting Mixture Distributions
• DiffusionBlocks: Block-wise Neural Network Training via Diffusion Interpretation
• Optimal structure learning and conditional independence testing
• ERUPT: An Open Toolkit for Interfacing with Robot Motion Planners in Extended Reality
• SIMSplat: Predictive Driving Scene Editing with Language-aligned 4D Gaussian Splatting
• U-LAG: Uncertainty-Aware, Lag-Adaptive Goal Retargeting for Robotic Manipulation
• A fast non-reversible sampler for Bayesian finite mixture models
• Discrimination in machine learning algorithms
• Post Reinforcement Learning Inference
• Extending Mean-Field Variational Inference via Entropic Regularization: Theory and Computation
• Fractional signature: a generalisation of the signature inspired by fractional calculus
• Statistical Inference for Temporal Difference Learning with Linear Function Approximation
• Iteratively reweighted kernel machines efficiently learn sparse functions
• Highly Efficient and Effective LLMs with Multi-Boolean Architectures
• A Nearly Optimal and Low-Switching Algorithm for Reinforcement Learning with General Function Approximation
• Variance Reduction and Low Sample Complexity in Stochastic Optimization via Proximal Point Method
• Batched Nonparametric Contextual Bandits
• On diffusion posterior sampling via sequential Monte Carlo for zero-shot scaffolding of protein motifs
• A semiconcavity approach to stability of entropic plans and exponential convergence of Sinkhorn's algorithm
• Orthogonal Procrustes problem preserves correlations in synthetic data
• Learning a distance measure from the information-estimation geometry of data
• What is in the model? A Comparison of variable selection criteria and model search approaches
• VisitHGNN: Heterogeneous Graph Neural Networks for Modeling Point-of-Interest Visit Patterns
• Hyperparameter Loss Surfaces Are Simple Near their Optima
• Oracle-based Uniform Sampling from Convex Bodies
• Differentially Private Wasserstein Barycenters
• Gradient-enhanced global sensitivity analysis with Poincar{\'e} chaos expansions
• Distilled Protein Backbone Generation
• Rates of Convergence of Generalised Variational Inference Posteriors under Prior Misspecification
• Total Robustness in Bayesian Nonlinear Regression for Measurement Error Problems under Model Misspecification
• Why Do We Need Warm-up? A Theoretical Perspective
• Best-of-Majority: Minimax-Optimal Strategy for Pass@$k$ Inference Scaling
• Higher-arity PAC learning, VC dimension and packing lemma
• Predictive inference for time series: why is split conformal effective despite temporal dependence?
• Beyond Linear Diffusions: Improved Representations for Rare Conditional Generative Modeling
• Adaptive randomized pivoting and volume sampling
• Learning Multi-Index Models with Hyper-Kernel Ridge Regression
• Neural Jump ODEs as Generative Models