Papers
Tue 2 Jun 2026
647 papers
- WMLLM: Self-Evolving Optimization Agents via Predict-Then-Act World ModelingWorld models → World models for language agents +2
- Vision-Language Models are Fragile Multilingual AssociatorsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- Why It Hurts: Identifying the Drivers of Negative Thoughts in Emotional Support ConversationsLLM reasoning → Audio-language model reasoning
- ConnectED: A Curriculum-Aligned AI System for Vietnamese Instructional Lesson Planning and Student LearningLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
- Kernel Forge: An Agent Harness for LLM-based Generation and Optimization of CUDA KernelsLLM agents and tool use → Software engineering coding agents
- SCAIR: Schema-Conditioned Agentic Iterative Reasoning for Enterprise Knowledge GraphsRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG +1
- Expectation Alignment of Language Models for Real-World User ExpectationsLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
- Adaptive Retrieval Strategies for Biomedical Question AnsweringRetrieval-augmented generation → Medical and biomedical RAG
- Agentic Safety is an Epistemic Property, Not a Behavioral OneLLM agents and tool use → Agent threats, guardrails and governance
- MAGNIFIED: RL Fine-tuning of Multimodal Large Language Models for Motion PlanningMultimodal LLMs / vision-language models → Driving VLMs and open-vocabulary detection
- An LLM-Explainable DRL Framework for Passenger-Directed Autonomous DrivingEnd-to-end autonomous driving → Large language models for driving
- RIZZ: Routing Interactions to Near Zero-Interference Zones for Continual Adaptation of Black-Box AgentsLLM agents and tool use → Agent memory systems
- SkillHarness: Harnessing Safe Skills for Computer-Use AgentsLLM agents and tool use → Agent threats, guardrails and governance
- GridVQA-X: A Framework for Evaluating Multimodal Explainability MethodsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- Steady-Forcing: Balancing Spatial Persistence and Motion Continuity in Long-Horizon Nature Video DiffusionDiffusion models → Video generation
- Rotate2Think: Geometric Priming via Orthogonal Rotation to Improve Language Model ReasoningLLM reasoning → Reasoning benchmarks and evaluation
- SD-GRPO: Verifiable Segment Decomposition for Long-Form Vision-Language GenerationLLM reasoning → Multimodal reasoning +1
- A Dataset for Dynamic Human Preferences for Vision Language ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- ViMax: Agentic Video GenerationLLM agents and tool use → Multimodal and embodied agents
- Steer Where It Matters: Token-Level Visual-Sensitivity Steering for LVLMs Hallucination MitigationMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs +1
- FineGen: A VLM-based Multi-Agent Framework for Fine-Grained Image-Text Dataset ConstructionLLM agents and tool use → Multimodal and embodied agents
- CrowdMath: A Dataset of Crowdsourced Mathematical Research DiscussionsLLM reasoning → Mathematical reasoning
- Agentic Large Language Models for Automated Structural Analysis of 3D Frame SystemsLLM reasoning → Spatial and embodied reasoning +1
- Lean4Agent: Formal Modeling and Verification for Agent Workflow and TrajectoryLLM agents and tool use → Software engineering coding agents
- SafeGene: Reusable Adapters for Transferable Safety AlignmentLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
- DiBS: Diffusion-Informed Branch SelectionDiffusion models → Discrete and text diffusion
- MAOAM: Unified Object and Material Selection with Vision-Language ModelsMultimodal LLMs / vision-language models → Image editing and quality assessment
- The Saturation Trap and the Subjectivity of Intervention Timing: Why Affect-Based Triggers and LLM Judges Fail to Time Interventions on Autonomous AgentsLLM agents and tool use → Agent threats, guardrails and governance
- Sparse Mixture-of-Experts Reward Models Learn Interpretable and Specialized Experts for Personalized Preference ModelingLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
- FindIt: A Format-Informed Visual Detection Benchmark for Generalist Multimodal LLMsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- From Ticks to Flows: Dynamics of Neural Reinforcement Learning in Continuous EnvironmentsDeep reinforcement learning methods → RL theory and sample complexity
- StandardE2E: A Unified Framework for End-to-End Autonomous Driving DatasetsEnd-to-end autonomous driving → End-to-end system architecture
- Instant-Fold: In-Context Imitation Learning for Deformable Object ManipulationOffline RL and imitation learning from data → Robot manipulation imitation learning
- UniCanvas: A Diffusion-base Unified Model for Text-in-Image Joint GenerationMultimodal LLMs / vision-language models → Image editing and quality assessment +1
- StepPRM-RTL: Stepwise Process-Reward Guided LLM Fine-Tuning for Enhanced RTL SynthesisLLM reasoning → Reasoning for agents, planning and code
- VAMPS: Visual-Assisted Mathematical Problem Solving BenchmarkMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- Overview of the EReL@MIR 2025 Multimodal Document Retrieval Challenge (Track 1)Retrieval-augmented generation → Multimodal RAG
- Recover-LoRA for Aggressive Quantization: Reclaiming Accuracy in 2-Bit Language Models via Low-Rank Adaptation with Knowledge Distillation on Synthetic DataSynthetic data and self-improvement for LLMs → Code and instruction data synthesis +1
- Supportive Token Revealing for Fast Diffusion Language Model DecodingDiffusion models → Discrete and text diffusion
- MM-BizRAG: Rethinking Multimodal Retrieval-Augmented Generation for General Purpose Enterprise Q&ALLM reasoning → Multimodal reasoning +1
- Consensus is Strategically Insufficient: Reasoning-Trace Disagreement as a Knowledge-Representation SignalLLM reasoning → Reasoning for agents, planning and code
- SMAC-Talk: A Natural Language Extension of the StarCraft Multi-Agent Challenge for Large Language ModelsLLM agents and tool use → General LLM agent frameworks and benchmarks
- Exploring the Topology and Memory of Consensus: How LLM Agents Agree, Fragment, or Settle When Forming ConventionsLLM agents and tool use → General LLM agent frameworks and benchmarks
- Dual Advantage FieldsDeep reinforcement learning methods → Deep RL algorithms and representations
- GroupToM-Bench: Benchmarking Group Theory of Mind and Nonlinear Social Emergence in MLLMsWorld models → World models for language agents
- Exact Unlearning in Reinforcement LearningLLM safety and security → Machine unlearning +1
- KODA: Contrastive Representation Comparison and Alignment for Vision-Language Foundation ModelsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- When Autoregressive Consistency Hurts Safety AlignmentLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
- Smart Transportation Without Neurons -- Fair Metro Network Expansion with Tabular Reinforcement LearningDeep reinforcement learning methods → RL for autonomous driving and traffic
- CaloTrilogy: Toward a Breakthrough in One-Step, End-to-End, Physics-Guided Shower Generation for Modern CalorimetersDiffusion models → Flow matching, time series and science
- Decomposing Refusal Steering in Mixture-of-Experts ModelsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +2
- SocialCoach: Personalized Social Skill Learning with Agentic Tutoring and PracticeLLM agents and tool use → Educational agents and math problem solving
- Thinking Through Signs: PEEL as a Semiotic Scaffolding for Epistemically Accountable AI-Enabled ResearchLLM reasoning → Reasoning benchmarks and evaluation
- EvalStop: Using World Feedback to Detect and Correct Reward Overoptimization in Multi-Tenant RLHF PlatformsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
- Caught in the Act(ivation): Toward Pre-Output and Multi-Turn Detection of Credential Exfiltration by LLM AgentsLLM agents and tool use → Agent threats, guardrails and governance +1
- Stationarity-Aware Retrieval-Augmented Time Series ForecastingRetrieval-augmented generation → Time series and agricultural RAG
- CLAW: Learning Continuous Latent Action World Models via Adversarial Latent RegularizationWorld models → Embodied and robot world models +1
- SaliMory: Orchestrating Cognitive Memory for Conversational AgentsLLM agents and tool use → Agent memory systems
- Reflection Separation from a Single Image via Joint Latent DiffusionDiffusion models → Inverse problems and image restoration
- UltraEP: Unleash MoE Training and Inference on Rack-Scale Nodes with Near-Optimal Load BalancingMixture-of-experts → Efficient MoE training and inference systems
- Optimal Transport Flow Matching by DesignDiffusion models → Flow matching, time series and science
- Token Budgets: An Empirical Catalog of 63 LLM-Agent Budget-Overrun Incidents, with an Affine-Typed Rust Mitigation as a Case StudyLLM agents and tool use → Software engineering coding agents
- A Goal-Set Characterization of Task Composition in the Boolean Task AlgebraDeep reinforcement learning methods → Interpretable and logic-guided RL
- RUBAS: Rubric-Based Reinforcement Learning for Agent SafetyLLM agents and tool use → Agent threats, guardrails and governance +1
- Dive into the Scene: Breaking the Perceptual Bottleneck in Vision-Language Decision Making via Focus Plan GenerationVision-language-action models → World model-based VLA
- Toward Pre-Deployment Assurance for Enterprise AI Agents: Ontology-Grounded Simulation and Trust CertificationLLM agents and tool use → Agent threats, guardrails and governance
- Self-Distilled Policy GradientKnowledge distillation → On-policy and self-distillation for LLMs +1
- Imaginative Perception Tokens Enhance Spatial Reasoning in Multimodal Language ModelsLLM reasoning → Spatial and embodied reasoning +1
- Skill-RM: Unifying Heterogeneous Evaluation Criteria via Agent SkillLLM agents and tool use → Reinforcement learning for LLM agents +1
- Video-Mirai: Autoregressive Video Diffusion Models Need ForesightDiffusion models → Video generation
- Quantifying Faithful Confidence Expression in Large Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking
- Agentic Chain-of-Thought Steering for Efficient and Controllable LLM ReasoningLLM reasoning → Test-time compute and efficient thinking +1
- AgenticRL: Agentic Reinforcement Learning with Self-Refinement for Complex UAV NavigationLLM agents and tool use → Reinforcement learning for LLM agents
- Using Reward Uncertainty to Induce Diverse Behaviour in Reinforcement LearningLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
- Preference-Calibrated Human-in-the-Loop Reinforcement Learning for Robotic ManipulationDeep reinforcement learning methods → RL for robot locomotion and manipulation
- PointAction: 3D Points as Universal Action Representations for Robot ControlDiffusion models → Diffusion and flow policies for robotics
- Entropy Is Not Enough: Unlocking Effective Reinforcement Learning for Visual Reasoning via Vision-Anchored Token SelectionLLM reasoning → Multimodal reasoning
- Value-Aware Stochastic KV Cache Eviction for Reasoning ModelsLLM reasoning → Test-time compute and efficient thinking +1
- Knowledge Editing for Masked Diffusion Language ModelsDiffusion models → Discrete and text diffusion
- PatchScene: Patch-based Voxel Diffusion for Large-Scale Scene CompletionWorld models → World modeling from LiDAR generation
- Bootstrap Your Generator: Unpaired Visual Editing with Flow MatchingDiffusion models → Image generation, text-to-image and editing
- NetKV: Network-Aware Decode Instance Selection for Disaggregated LLM InferenceEfficient LLM inference and serving → Serving systems and request scheduling
- The Impact of Configuring Agentic AI Coding Tools on Build-vs-Buy Decisions: A Study ProtocolLLM agents and tool use → Software engineering coding agents
- Agent libOS: A Runtime Substrate for Capability-Controlled Self-Evolving LLM AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks +1
- Synthesize and Reward -- Reinforcement Learning for Multi-Step Tool Use in Live EnvironmentsLLM agents and tool use → Reinforcement learning for LLM agents
- OVO-S-Bench: A Hierarchical Benchmark for Streaming Spatial Intelligence in Multimodal LLMsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
- CoralBay: A Self-Supervised CT Foundation ModelKnowledge distillation → Distillation for medical imaging
- Reasoning Structure of Large Language ModelsLLM reasoning → Test-time compute and efficient thinking
- Unified Video-Action Joint Denoising for Dexterous Action and Data GenerationWorld models → Embodied and robot world models
- A Training-Free Mixture-of-Agents Framework for Multi-Document Summarization using LLMs and Knowledge GraphsLLM agents and tool use → Web, search, and research agents
- PyraMathBench: Evaluating and Improving Mathematical Capability in Large Language ModelsLLM reasoning → Mathematical reasoning
- CLI-Anything: Towards Agent-Native Computer UseLLM reasoning → Multimodal reasoning +1
- EvoDS: Self-Evolving Autonomous Data Science Agent with Skill Learning and Context ManagementLLM agents and tool use → General LLM agent frameworks and benchmarks +1
- Let the Dynamics Flow: Stable Flow Matching Dynamical SystemsDiffusion models → Diffusion and flow policies for robotics +1
- BigFinanceBench: A Workflow-Grounded Benchmark for Financial-Research AgentsLLM agents and tool use → Financial trading and forecasting agents
- Conditional Latent Diffusion Model with Fourier-based Motion Modelling for Virtual Population SynthesisDiffusion models → Medical imaging and segmentation
- A Quantitative Approximation Framework for Flow Distillation in Diffusion ModelsDiffusion models → Diffusion foundations and sampling
- TreeFlash: Parallel AR-Approximation for Faster Speculative DecodingEfficient LLM inference and serving → Speculative decoding
- Rethinking the Idiomaticity Decomposability Hypothesis: Evidence from Distributional LearningLLM reasoning → Reasoning benchmarks and evaluation
- Enhancing Operational Safety via Agentic Dialogue Hazard Identification AnalysisLLM agents and tool use → Agent threats, guardrails and governance
All 647
- AI Agents Enable Adaptive Computer WormsLLM agents and tool use → Agent threats, guardrails and governance
- PURGE: Projected Unlearning via Retain-Guided ErasureLLM safety and security → Machine unlearning
- Easy-to-Use Shielding for Reinforcement LearningDeep reinforcement learning methods → Safe and constrained RL
- Trading Human Curation for Synthetic Augmentation in RLVRLLM agents and tool use → Reinforcement learning for LLM agents
- Beyond Compression: Quantifying Spectral Accessibility in Vision RepresentationsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- Exploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness +2
- Training-Free Multi-Concept LoRA Composition with Prompt-Aware WeightingDiffusion models → Image generation, text-to-image and editing
- Backdoor Unlearning Generalization: A Path Toward the Removal of Unknown Triggers in LLMsLLM safety and security → Backdoor and data poisoning attacks
- Revisiting Embodied Chain-of-Thought for Generalizable Robot ManipulationLLM reasoning → Spatial and embodied reasoning +1
- Reasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?LLM reasoning → Reasoning benchmarks and evaluation
- Expert-Aware Causal Tracing of Factual Recall in Sparse MoE Language ModelsMixture-of-experts → Expert routing strategies +1
- From Control Boundary to Insurance Claim: Reconstructing AI-Mediated Losses Through the CER FrameworkLLM agents and tool use → Financial trading and forecasting agents
- E2LLM: Towards Efficient LLM Serving in Heterogeneous Edge/Fog EnvironmentsEfficient LLM inference and serving → Serving systems and request scheduling
- HybridThinker: Efficient Chain-of-Thought Reasoning via Compressed Memory and Transient Thought StepsLLM reasoning → Test-time compute and efficient thinking
- Tool-Aware Optimization with Entropy Guidance for Efficient Agentic Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents
- LAP: An Agent-to-Instrument Protocol for Autonomous ScienceLLM agents and tool use → Web, search, and research agents
- Entropy Gate: Entropy Quenching for Near-Lossless Token Compression in LLM PipelinesEfficient LLM inference and serving → KV cache optimization
- Beyond False Stability: High-Noise Drift Gating for Test-Time Adversarial Defenses in Vision-Language ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
- Re-Ranking Through an Attribution Lens for Citation Quality in Legal QARetrieval-augmented generation → RAG robustness and adaptive retrieval
- Investigating Adversarial Robustness of Multi-modal Large Language ModelsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
- Code-on-Graph: Iterative Programmatic Reasoning via Large Language Models on Knowledge GraphsLLM reasoning → Knowledge- and tool-augmented reasoning
- Multi2: Hierarchical Multi-Agent Decision-Making with LLM-Based Agents in Interactive EnvironmentsLLM agents and tool use → Reinforcement learning for LLM agents
- Does Language Shift Break Medical Vision-Language Models? Indonesian Radiology Visual Question Answering Case StudyMultimodal LLMs / vision-language models → Medical vision-language models
- The DeepSpeak-Agentic DatasetLLM agents and tool use → General LLM agent frameworks and benchmarks
- A Close Look At World Model Recovery In Supervised Fine-Tuned LLM PlannersWorld models → World models for language agents
- EvoDrive: Pareto Evolution for Safety-Critical Autonomous Driving via Self-Improving LLM AgentsLLM agents and tool use → Multimodal and embodied agents +1
- Agentic Generation and Evolution of Knowledge ModelsLLM agents and tool use → Software engineering coding agents
- From Answers to States: Verifiable Process-Level Evaluation of Chemical Reasoning in Large Language ModelsLLM reasoning → Clinical and biomedical applications
- Diagnosing Knowledge Gaps in LLM Tool Use: An Agentic Benchmark for Novel API AcquisitionLLM agents and tool use → Software engineering coding agents
- All you need to break LLMs are Black-Box, Adapting, Efficient, Transferable, Harmful, Applicable ... AttacksLLM safety and security → Jailbreaking
- AnchorMoE: Interpretable Time Series Classification via Anchor-Routed MoEMixture-of-experts → MoE for time series and graphs
- TSQAgent: Rating Time Series Data Quality via Dedicated Agentic ReasoningLLM reasoning → Time series reasoning +1
- TurtleAI: Benchmarking Multimodal Models for Visual Programming in Turtle GraphicsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
- Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning ModelsLLM reasoning → Reasoning benchmarks and evaluation
- Physics-Guided Policy Optimization with Self-DistillationKnowledge distillation → On-policy and self-distillation for LLMs
- Cross-Lingual Token Arbitrage: Optimizing Code Agent Context Windows via Local LLM PreprocessingLLM agents and tool use → General LLM agent frameworks and benchmarks
- A 3D Isovist World Model -- Revealing a City's Unseen Geometry and Its Emergent Cross-City SignatureWorld models → World models for navigation
- Exploiting Verification-Generation Gap: Test-Time Reinforcement Learning with Confidence-Conditioned VerificationLLM reasoning → Reinforcement learning for reasoning +1
- Testing LLM Arithmetic Reasoning Generalization with Automatic Numeric-Remapping AttacksLLM reasoning → Mathematical reasoning
- Beyond the Literal: Decomposing Pragmatic Intent in Multimodal Meme UnderstandingMultimodal LLMs / vision-language models → MLLM training and visual instruction tuning
- World Models Meet Language Models: On the Complementarity of Concrete and Abstract ReasoningWorld models → World models for language agents +2
- DDOR: Delta Debugging for Explainable Overrefusal Testing and RepairLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
- PHASER: Phase-Aware and Semantic Experience Replay for Vision-Language-Action ModelsVision-language-action models → Reinforcement learning and post-training
- Diffusing in the Right Space: A Systematic Study of Latent DiffusabilityDiffusion models → Diffusion foundations and sampling
- Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline MatchingLLM reasoning → Spatial and embodied reasoning +1
- When Attention Collapses: Stage-Aware Visual Token Pruning from Structure to SemanticsMultimodal LLMs / vision-language models → Visual token pruning and efficiency +1
- Skill Is Not Document: Query-Conditioned Compatibility for LLM Agent Skill RoutingLLM agents and tool use → Agent skill learning and evolution
- CR-Seg: Attention-Guided and CoT-Enhanced Coarse-to-Refined Reasoning SegmentationLLM reasoning → Multimodal reasoning
- Partially Observable Adversarial Patch Attacks on Vision-Language-Action Models in RoboticsVision-language-action models → Safety and adversarial attacks
- When Should the Teacher Move? Temporal Coupling and Stability in Self On-Policy DistillationKnowledge distillation → On-policy and self-distillation for LLMs
- Post-Hoc Robustness for Model-Based Reinforcement LearningDeep reinforcement learning methods → Adversarial robustness and RL in finance
- Overlaying Governance: A Compositional Authorization Framework for Delegation and Scope in Agentic AILLM agents and tool use → Agent threats, guardrails and governance
- SPADE: Sketch-guided Path Planning Augmented with Diffusion ExpertsOffline RL and imitation learning from data → Imitation and offline RL for driving
- EvoMemNav: Efficient Self-Evolving Fine-Grained Memory for Zero-Shot Embodied NavigationSynthetic data and self-improvement for LLMs → Self-improving MLLMs and synthetic data
- ThoughtFold: Folding Reasoning Chains via Introspective Preference LearningLLM reasoning → Test-time compute and efficient thinking
- Characterizing Detectability in 3DGS Poisoning: A Stage-wise BenchmarkLLM safety and security → Backdoor and data poisoning attacks
- Learn from Your Mistakes: Tree-like Self-Play for Secure Code LLMsSynthetic data and self-improvement for LLMs → Code and instruction data synthesis
- NeuroArmor: Safe-Variant-Guided Representation Consistency for Selective Re-Anchoring in Jailbreak DefenseLLM safety and security → Jailbreaking
- Human2Humanoid: Physics-Aware Cross-Morphology Motion Retargeting for Humanoid RobotsOffline RL and imitation learning from data → Real-world robot offline RL
- StepFinder: A Temporal Semantic Framework for Failure Attribution in Multi-Agent SystemsLLM agents and tool use → Software engineering coding agents
- Rethinking the Role of Tensor Decompositions in Post-Training LLM CompressionLLM reasoning → Test-time compute and efficient thinking
- DMF: A Deterministic Memory Framework for Conversational AI AgentsLLM agents and tool use → Agent memory systems
- KVarN: Variance-Normalized KV-Cache Quantization Mitigates Error Accumulation in Reasoning TasksLLM reasoning → Test-time compute and efficient thinking +1
- WavTTS: Towards High-Quality Zero-Shot TTS via Direct Raw Waveform ModelingDiffusion models → Audio, speech and music
- FORGE: Multi-Agent Graduated Exploitation and Detection EngineeringLLM agents and tool use → Agent threats, guardrails and governance
- PRISM: Synergizing Vision Foundation Models via Self-organized Expert SpecializationMixture-of-experts → Vision mixture-of-experts
- CP-Agent: Context-Aware Multimodal Reasoning for Cellular Morphological Profiling under Chemical PerturbationsLLM reasoning → Clinical and biomedical applications +1
- MeDxAgent: Multi-Agent Consultation for Interactive Medical DiagnosisLLM agents and tool use → Medical and scientific domain agents
- Enginuity: A Dataset and Benchmark for Vision-Language Understanding of Engineering DiagramsMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
- Mamba-Enhanced Implicit Motion Learning for Audio-Driven Portrait AnimationState-space sequence models (S4, Mamba) → Video, motion and generative Mamba models
- Human-AI Collaboration and the Transformation of Software Engineering WorkLLM agents and tool use → Software engineering coding agents
- Flicker-DDPM: Accelerating Denoising Diffusion via 1/f Colored Noise InjectionDiffusion models → Diffusion foundations and sampling
- When Model Merging Breaks Routing: Training-Free Calibration for MoEMixture-of-experts → Expert routing strategies
- Local Guidance, Global Impact: Gaussian-Reshaped Trust Region Unlocks Behavior TransitionsDeep reinforcement learning methods → RL theory and sample complexity
- P2-DPO: Grounding Hallucination in Perceptual Processing via Calibration Direct Preference OptimizationMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs +1
- eMEM: A Hybrid Spatio-Temporal Memory System For Embodied AgentsLLM agents and tool use → Multimodal and embodied agents
- See Better, Foresee Better, Act Wiser: Physically Grounded Proactive Modeling and Decision MakingWorld models → Embodied and robot world models
- Mitigating False Credit Propagation: Probabilistic Graphical Reward Aggregation for Rubric-Based Reinforcement LearningDeep reinforcement learning methods → RL training of language and agent models
- ImageAuditor: Membership Inference Attack against Image-based Retrieval-Augmented GenerationRetrieval-augmented generation → Privacy and security of RAG +1
- AugMask: Training Diffusion Models on Incomplete Tabular Data via Stochastic Augmentation and MaskingDiffusion models → Flow matching, time series and science
- Cross-Modality Feature Fusion Based on Structured State Space Duality for Multimodal Image Registration NetworkState-space sequence models (S4, Mamba) → Point clouds, detection and remote sensing
- A GPU-Parallel Framework for Heterogeneous Multi-Task Reinforcement LearningDeep reinforcement learning methods → RL for robot locomotion and manipulation
- InfoMem: Training Long-Context Memory Agents with Answer-Conditioned Information GainLLM agents and tool use → Agent memory systems
- CAPER: Clause-Aligned Process Supervision for Text-to-SQLLLM alignment / RLHF / preference optimisation → Process reward models for reasoning
- Generalizing Graph Foundation Models via Hyperbolic Retrieval-Augmented GenerationRetrieval-augmented generation → Knowledge-graph and structured-knowledge RAG
- LEAP: Supercharging LLMs for Formal Mathematics with Agentic FrameworksLLM reasoning → Mathematical reasoning +1
- SagaQA: A Multi-hop Reasoning Benchmark for Long-form Narrative Understanding in TV SeriesLLM reasoning → Multimodal reasoning
- SplitAdapter: Load-Aware Humanoid Loco-Manipulation via Factorized AdaptationWorld models → Humanoid and legged locomotion world models
- Bridging Predictive Uncertainty and Safe Action: Sample-Conditioned Differentiable Planning for Autonomous DrivingEnd-to-end autonomous driving → Imitation-learning end-to-end planners
- Multilingual Unlearning in LLMs: Transfer, Dynamics, and ReversibilityLLM safety and security → Machine unlearning
- VistaHop: Benchmarking Long-Horizon Visual DeepSearchMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
- Agentic Relationship Harm: Benchmarking and Gating Relational Manipulation in AI AgentsLLM agents and tool use → Agent threats, guardrails and governance
- EaDex: A Cross-Embodiment Dexterous Manipulation Framework from Low-Cost DemonstrationsOffline RL and imitation learning from data → Robot manipulation imitation learning
- AirDreamer: Generalist Drone Navigation with World ModelsWorld models → World models for navigation
- GeoAlign: Beyond Semantics with State-Guided Spatial Alignment in VLA ModelsVision-language-action models → 3D and spatial-aware VLA
- ARBOR: Online Process Rewards via a Reusable Rubric Buffer for Search AgentsLLM agents and tool use → Web, search, and research agents
- When RLHF Fails: A Mechanistic Taxonomy of Reward Hacking, Collapse, and Evaluator GamingLLM alignment / RLHF / preference optimisation → Reward hacking and misspecification
- Perceive Before Reasoning: A Pre-Reasoning Perception Framework for Efficient and Reliable Proactive Mobile AgentsLLM reasoning → Multimodal reasoning +2
- Right Makes Might: Aligning Verified Hidden States Empowers RL ReasoningLLM reasoning → Reinforcement learning for reasoning
- An Asymptotic Theory of Chain-of-Thought in In-Context LearningLLM reasoning → Prompting and chain-of-thought reasoning
- Follow-Your-Preference++: Rethinking Preference Alignment for Image InpaintingLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
- Bayesian Tensor Decomposition with Diffusion Model PriorDiffusion models → Inverse problems and image restoration
- MedCUA-Bench: A Screenshot-Only Benchmark for Clinical Computer-Use AgentsLLM agents and tool use → Medical and scientific domain agents
- Reinforcement Learning from Cross-domain Videos with Video Prediction ModelDeep reinforcement learning methods → RL for robot locomotion and manipulation
- MemTrain: Self-Supervised Context Memory TrainingLLM agents and tool use → Agent memory systems
- Focused on the User, Overlooking the Risks: Security and Privacy Understandings, Practices and Challenges of Independent Chinese AI Agent DevelopersLLM agents and tool use → Agent threats, guardrails and governance
- GeoSem-WAM: Geometry- and Semantic-Aware World Action ModelsWorld models → Embodied and robot world models
- Inference-Time Scaling for Joint Audio-Video GenerationLLM reasoning → Audio-language model reasoning
- GLINT: Sparsely Gated Vision-Language Alignment for Fine-Grained Radiology RepresentationsMultimodal LLMs / vision-language models → Medical vision-language models +1
- Fully Automated Identification of Lexical Alignment and Preference-Stage Shifts in Large Language ModelsLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
- NVIDIA OmniDreams: Real-Time Generative World Model for Closed-Loop Autonomous Vehicle SimulationWorld models → World models for autonomous driving +1
- GTBench: A Curriculum-Grounded Benchmark for Evaluating LLMs as Mathematical Research Assistants in Graph TheoryLLM reasoning → Mathematical reasoning
- FederatedSkill: Federated Learning for Agentic Skill EvolutionLLM agents and tool use → Agentic AI for wireless networks
- Disentangling Visual and Factual Correctness in LVLMs' Visualization LiteracyMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
- Think-Before-Speak: From Internal Evaluation to Public Expression in Multi-Agent Social SimulationLLM reasoning → Reasoning for agents, planning and code +1
- PsychoPass: Geometric Profiling of Multi-Turn Adversarial LLM ConversationsLLM safety and security → Jailbreaking
- Uncertainty-Aware Clarification in LLM Agents with Information GainLLM agents and tool use → General LLM agent frameworks and benchmarks
- HARVE: Hacking-Aware Reward-Head Vector Editing for Robust Reward ModelsLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
- TTT-VLA: Test-Time Latent Prompt Optimization for Vision-Language-Action ModelsVision-language-action models → Reinforcement learning and post-training
- GuidedBridge: Training-freely Improving Bridge Models with Prior GuidanceDiffusion models → Diffusion foundations and sampling
- Inverting the Generation Process of Denoising Diffusion Implicit Models: Empirical Evaluation and a Novel MethodDiffusion models → Inverse problems and image restoration
- Coherence Maximization Improves Pluralistic AlignmentLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
- EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement LearningLLM agents and tool use → Reinforcement learning for LLM agents
- DeskCraft: Benchmarking Desktop Agents on Professional Workflows and Human-in-the-Loop CollaborationLLM agents and tool use → GUI and mobile agents
- Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time ScalingLLM reasoning → Reinforcement learning for reasoning
- Zero-Shot 3D Question Answering via Hierarchical View-to-Token TransportationLLM reasoning → Spatial and embodied reasoning +1
- PhotoCraft: Agentic Reasoning with Hierarchical Self-Evolving Memory for Deep Image SearchLLM reasoning → Multimodal reasoning +2
- FGRPO: Federated GRPO with Adaptive Aggregation on Non-IID DataLLM reasoning → Reinforcement learning for reasoning
- Decomposing how prompting steers behaviorMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
- The Shadow Price of Reasoning: Economic Perspective on Optimal Budget Allocation for LLMsLLM reasoning → Test-time compute and efficient thinking
- "**Important** You should give me full credits!": Exploring Prompt Injection Attacks on LLM-Based Automatic Grading SystemsLLM safety and security → Prompt injection and agent security
- Constitutional On-Policy Safe DistillationKnowledge distillation → On-policy and self-distillation for LLMs
- DELTAMEM: Incremental Experience Memory for LLM Agents via Residual TreesLLM agents and tool use → Agent memory systems
- Libra: Efficient Resource Management for Agentic RL Post-TrainingLLM agents and tool use → Reinforcement learning for LLM agents
- TGV-KV: Text-Grounded KV Eviction for Vision-Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
- RMPrior: Bridging Propagation Priors and Diffusion Refinement for Efficient Radio Map ConstructionDiffusion models → Diffusion for compression and wireless
- CORE: Conflict-Oriented Reasoning for General Multimodal Manipulation DetectionLLM reasoning → Multimodal reasoning
- Rethinking Molecular Text Representations for LLMs: An Empirical StudyLLM reasoning → Clinical and biomedical applications
- SkillDAG: Self-Evolving Typed Skill Graphs for LLM Skill Selection at ScaleLLM agents and tool use → Agent skill learning and evolution +1
- ToolGate: Token-Efficient Pre-Call Control for Tool-Augmented Vision-Language AgentsLLM agents and tool use → Multimodal and embodied agents
- What Do Students Learn? A Feature-Level Analysis of Dark KnowledgeKnowledge distillation → General knowledge distillation methods
- Capability Advertisement as a Market for Lemons: A Trust Layer for Heterogeneous Agent NetworksLLM agents and tool use → Financial trading and forecasting agents
- The Deliberative Illusion: Diagnosing Factual Attrition and Stance Homogenization in Multi-Agent LLM DeliberationLLM agents and tool use → Social simulation and multi-agent debate
- AUDITFLOW: Executable Symbolic Environments for Structured Financial Reporting VerificationLLM agents and tool use → Financial trading and forecasting agents
- SkillGuard: A Permission-Centric Framework for Agent Skill SecurityLLM agents and tool use → Agent threats, guardrails and governance
- Hint-Guided Diversified Policy Optimization for LLM ReasoningLLM reasoning → Reinforcement learning for reasoning
- ConTraIRL: Factorized Contrastive Abstractions for Transferable IRLOffline RL and imitation learning from data → Inverse reinforcement learning +1
- MOSAIC: Efficient Mixture-of-Agent Scheduling via Adaptive Aggregation and Inference ConcurrencyLLM agents and tool use → Software engineering coding agents
- MUSE: A Unified Agentic Harness for MLLMsMultimodal LLMs / vision-language models → Embodied and 3D vision-language models +1
- Exact equivariance, kept through training, buys zero-shot generalisation across the symmetry groupWorld models → Latent predictive world models
- Perplexity Can Miss SAE Feature Damage Under QuantizationMechanistic interpretability → SAE features and their applications
- Patcher: Post-Hoc Patching of Backdoored Large Language ModelsLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
- Inducing Reasoning Primitives from Agent TracesLLM reasoning → Reasoning for agents, planning and code +1
- Predicting Inference-Time Scaling Gains from Labeled Validation-Set Output StatisticsLLM reasoning → Test-time compute and efficient thinking +1
- Transformer Models for Text Summarization: A Comparative Study of BART, BERT, and RoBERTa
- Position: Certified Correctness in Neural Constraint Reasoning Requires Symbolic Integration
- Steering the Language Axis: From Linear Decodability to Causal Control
- Can Spectral-Clipping Enable Better Learning While Forgetting Less for Low-Rank Adaptation?
- HyperShape: Hyperelasticity Across Diverse Shapes
- Unanticipated Effects of Generative AI on Expertise Pathways and Performance Perception in System Administration
- COSI-Lab: Conference Living Lab for Modeling Multi-Perspective Multimodal Social Intention
- Directional Hallucinations: Ideological Drift in News-Grounded LLM Question Answering
- OPTScientist: Multi-Agent Discovery of Typed Optimizer Programs for Transformer Pretraining
- The Devil is in the Spectrum: Mitigating Representation Collapse in LLMs via Topologically Regularized Side-Path
- Lifted State Hypothesis in Large Language Models
- Trajectory-Aware Clinical Risk Prediction via Severity-Grounded Knowledge Graphs and Retrieval-Augmented Generation
- Collaborative Disagreement Resolution for Scalable Oversight
- Structuring the Space of Sociotechnical Alignment
- A Practice Auditing Framework for Large Language Model Use: Collective Empiricism, Pseudo-Rational Cognition, and Governance of AI-Generated Content
- Auditing LLM-Governed Social Robots with Culture-Specific Moral Gradients
- Constituency Optimisation Through Hamiltonian Representation Of Mandates (COTHROM): Algorithmic Redistricting of Irish Election Boundaries
- HorusEye: Language as Dynamic Attention for Emergency Visual Analysis
- An RRAM-based Hardware Implementation of a Radial Basis Function Neuron for Edge Classifiers
- Learning Topological Representations of Protein Structure and Dynamics
- UtVAA: Ultra-tiny Vision Transformer with Affix Attention for Mobile Image Classification
- BRIDGE: Biological Evidence Refinement and Heterogeneous Dynamic Gating for Gene Regulatory Networks
- BBR-Net: Boundary-Balanced Replay for Continual Medical Image Segmentation
- Hierarchical GRU with Input-Conditioned Slot Queries for Ball Action Anticipation
- Calibrating Overconfidence Without Sacrificing Confidence: Probe-Conditioned Head Intervention for LLMs
- Integrating Local and Global Entropy for Uncertainty Quantification in LLMs
- Disjoint or Overlapping? Inference Windowing for Reconstruction-Based Time Series Anomaly Detection
- PatchSTG: Scalable Spatiotemporal Graph Transformers for Traffic Forecasting on Irregular Sensor Networks
- KITE: A Tri-Modal Transformer Integrating Text, Images, and Knowledge Graphs for Fake News Detection
- Detecting Aimbot Cheaters in MOGs
- AQIFormer: A Transformer-Based Multi-View Architecture for Cross-City Air Quality Classification
- DOME: Learning Transferable Domain Variables from Sparse Supervision for Test-Time Adaptation
- Advanced Flood Prediction with Physics-Guided Deep Learning: Combining UNet, FNO, and SAR/Optical Imagery
- P-Cast Precision in FP8 Attention: Sink-Induced Collapse and the Optimality of S=2^8
- Applying Deep Learning for cockpit segmentation in the context of mixed reality
- Probabilistic learning to perform pre-onset individualised prediction of disease severity: application to Veno Occlusive Disease
- DxPTA: An Architecture Design Space Exploration with Optical Dataflow-guided Strategy for HW/SW Co-Design of Photonic Transformer Accelerators
- Detecting and Mitigating Bias by Treating Fairness as a Symmetry Operation
- Central Description Length (CDL) Clustering Validation Index
- Quantifying the biophysical properties of stomatocytes in health and disease
- The Language of Elution: Autoregressive Prediction of the Next Feature in Untargeted LC-HRMS Lipidomics
- DiverAge: Reliable Pluralistic Face Aging with Cross-Age Identity Relation Guidance
- Anycast Performance in Context
- A Cookbook of 3D Vision: Data, Learning Paradigms, and Application
- PE-MHL: Physics-Encoded Modular Hybrid Layers for Scalable Learning of Complex Systems
- Scaling Novel Graph Generation via Lightweight Structure-Guided Autoregressive Models
- Using Text-Based Causal Inference to Disentangle Factors Influencing Online Review Ratings
- The Loss Is Not Enough: Sampling Conditions and Inductive Bias in Contrastive Representation Learning
- Derivative Informed Learning of Exchange-Correlation Functionals
- Long Live Fine-Tuning: Task-Specific Transformers Outperform Zero-Shot LLMs for Misinformation Response Classification on Reddit
- Human agency in initial human-AI proof formalization workflows
- RL Excursions during Pre-Training: Re-examining Policy Optimization for LLM training
- Long-Term and Short-Term Transistor Aging in Deep Neural Networks: Impact and Mitigation
- Nonlocal Mean Field Schr"odinger Bridge with Learned Interactions
- Can I Take Another Dose? Evaluating LLM Decision-Making Under Temporal Uncertainty in OTC Dosing QA
- Can Generalist Agents Automate Data Curation?
- SBP-Net: Learning Thin Structure Reconstruction with Sliding-Box Projections
- Prospective Dynamic 3D MRI Reconstruction via Latent-Space Motion Tracking from Single Measurement
- Incremental Sheaf Cohomology on Cellular Complexes: O(1)-in-n Lazy Edit Processing under Bounded Local Geometry
- PerceptTwin: Semantic Scene Reconstruction for Iterative LLM Planning and Verification
- Edge of Stability Selectively Shapes Learning Across the Data Distribution
- Representation Matters in Randomized Smoothing for Audio Classification
- A Geometric View of Counterfactual Behavior: Interaction of Boundary Proximity and Local Support
- DetectZoo: A Unified Toolkit for AI-Generated Content Detection Across Text, Audio, and Image Modalities
- Cross-Prompt Generalization in Detecting AI-Generated Fake News Using Interpretable Linguistic Features
- Spatial Artifact Coherence Determines Codec Robustness in Patch-Based rPPG
- Training-Free Lexical-Dense Fusion for Conversational-Memory Retrieval
- Notarized Agents: Receiver-Attested Confidential Receipts for AI Agent Actions
- Metric-Aware Hybrid Forecasting for the CTF4Science Lorenz Challenge
- ACAT: A Collaborative Platform for Efficient Aspect-Based Sentiment Dataset Annotation
- A Systematic Analysis of Linguistic Features in AI-Generated Text Detection Across Domains and Models
- Low-rank Distributional Matrix Completion
- MimeLens: Position-Agnostic Content-Type Detection for Binary Fragments
- End-to-End Text Line Detection and Ordering
- ADAPTOOD: Uncertainty-Aware Fine-Tuning for Out-of-Distribution ECG Time Series Models
- When Offline Selectors Cannot Beat the Best Single Model: A Diagnostic Study on edX Dropout Prediction
- EpiFormer: Learning Antigen-Antibody Interactions for Epitope Prediction via Geometric Deep Learning
- Stumbling Into AI Emotional Dependence: How Routine AI Interactions Reshape Human Connection
- Physics-Informed Machine Learning for Short-Term Flood Prediction
- Pinpoint: Grounded Worldwide Image Geolocation via Cross-Source Retrieval and Reranking
- When Retrieval Doesn't Help: A Large-Scale Study of Biomedical RAG
- HighTide: An Agent-Curated Open-Source VLSI Benchmark Suite
- Semantic Constraint Synthesis for Adaptive Trajectory Optimization via Large Language Models
- veriFIRE: an Industrial Case Study in Verifying Consistency Properties for a DNN-Based Wildfire Detection System
- Computational conceptual history of scientific concepts: From early digital methods to LLMs
- dMX: Differentiable Mixed-Precision Assignment for Low-Precision Floating-Point Formats
- AgenticDiffusion: Multi-View Reasoning with View-Conditioned Diffusion Planning for Vision-Based UAV Navigation
- Variance Reduction for Heavy-Tailed Monetization Metrics in Ranking Experiments via Post-Stratification
- Discourse-Role Labels as Presentation-Time Variables for Context Use in Language Models
- SymTRELLIS: Symmetry-Enforced Voxel Latents for 3D Generation
- Building The Ph(ysical)AI Layer Of Machine Intelligence
- Proof-Carrying Agent Actions: Model-Agnostic Runtime Governance for Heterogeneous Agent Systems
- The Differentiable Auditory Loop (DAL): An ML Framework for Hyper-Personalized Hearing Aids
- Stein Kernelized Molecular Dynamics for Active Learning of Interatomic Potentials
- When Seeing Is Not Believing -- A Benchmark for Search-Grounded Video Misinformation Detection
- POLARIS: Guiding Small Models to Write Long Stories
- Large Language Models Hack Rewards, and Society
- Adaptive Patching Is Harder Than It Looks For Time-Series Forecasting
- TPA-AD: A Two-Stage Pseudo Anomaly-Guided Method for Bearing Time-Series Anomaly Detection
- CADET: A Modular Platform for Evaluating Distributed Cooperative Autonomy in Connected Autonomous Vehicles
- Covert Influence Between Language Models
- Bayesian Membership Privacy for Graph Neural Networks
- Need to Know: Contextual-Integrity-Grounded Query Rewriting for Privacy-Conscious LLM Delegation
- SC-TauPath: A Structural Connectivity Attribution Framework for Mapping Tau Propagation Pathways in Alzheimer's Disease
- Finite-Iteration Local Dynamics and Warm Starts for Alternating Power Iteration in Spiked Tensor PCA
- LLM Compression with Jointly Optimizing Architectural and Quantization choices
- Intra-Modal Neighbors Never Lie: Rectifying Inter-Modal Noisy Correspondence via Graph-Based Intra-Modal Reasoning
- Weakly Supervised Incremental Segmentation via Semantic Anchors and Spatial Arbitration
- Spectral Scaling Laws of Muon
- The Invisible Lottery: How Subtle Cues Steer Algorithm Choice in LLM Code Generation
- LiftQuant: Continuous Bit-Width LLM via Dimensional Lifting and Projection
- Unlocking Feature Learning in Gated Delta Networks at Scale
- Bayes-Sufficient Representations in Supervised Learning
- Channel-Preserving Representation Alignment for EEG-to-Music Reconstruction
- Beyond Static Priors: Dynamic Neural Guidance for Large-Scale Ant Colony Optimization
- SimuScene: Simulation-Ready Compositional 3D Scene Reconstruction from a Single Image
- Exploring Easy Boosts for Lidar Semantic Scene Completion
- Neuron Populations Exhibit Divergent Selectivity with Scale
- PixVOD: Pixel-Distributed Direct Visual Odometry and Depth Estimation
- NewtPhys: Do Foundation Models Understand Newtonian Physics?
- Humanoid-GPT: Scaling Data and Structure for Zero-Shot Motion Tracking
- Language Models Compare Quantities Using Number-specific and Unit-specific Heuristics
- Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories
- Formalizing the Binding Problem
- AAD-1: Asymmetric Adversarial Distillation for One-Step Autoregressive Video Generation
- QUBRIC: Co-Designing Queries and Rubrics for RL Beyond Verifiable Rewards
- AlignAtt4LLM: Fast AlignAtt for Decoder-Only LLMs at IWSLT 2026 Simultaneous Speech Translation Task
- Efficient ASR Training with Conversations that Never Happened
- VLESA: Vision-Language Embodied Safety Agent for Human Activity Monitoring
- Demo2Tutorial: From Human Experience to Multimodal Software Tutorials
- A Pocket Offline Model for Simultaneous Speech Translation as CUNI Submission to IWSLT 2026
- MLSkip: Data Skipping for ML Filters via Lightweight Metadata
- SEAOTTER: Sensor Embedded Autoencoding with One-Time Transcode for Efficient Reconstruction
- FlashbackCL: Mitigating Temporal Forgetting in Federated Learning
- q0: Primitives for Hyper-Epoch Pretraining
- Correcting Neural Operator Spectral Bias via Diffusion Posterior Sampling with Sparse Observations
- Quadratic integrate-and-fire neurons exhibit less fragmented loss landscapes and outperform leaky integrate-and-fire neurons in spike-based gradient descent
- FFR: Forward-Forward Learning for Regression
- DiffUNet^2: Bidirectional Conditional Diffusion for Probabilistic Scientific Spatiotemporal Modeling
- Adaptive Causal Alignment for High-Confidence Adversarial Training
- Contrastive Neural Algorithmic Reasoning for Graph Coloring
- GARDEN: Gravity-Aligned Reconstruction of Disentangled ENvironments from RGB images
- Benchmarking Visual State Tracking in Multimodal Video Understanding
- Forecasting Conceptual Diffusion in Science: The Case of Quantum Computing
- Hedge-Bench: Benchmarking Agents on Hard, Realistic Tasks Pertaining to Financial Reasoning
- Beyond Gradient Descent: Adam for Analog Ising Machines
- SparseStreet: Sparse Gaussian Splatting for Real-Time Street Scene Simulation
- scTranslation: A Comprehensive Benchmark for Single-Cell Multi-Omics Modality Translation
- MAdam: Metric-Aware Multi-Objective Adam
- An Attention-Based Denoising Model for Diffusion Weighted Imaging
- Denoise First, Orthogonalize Later: Understanding Momentum in Muon via Spectral Filtering
- Electromagnetic Navigation for Femoral Osteotomy Using High-Accuracy X-ray-to-CT Registration
- RealClawBench: Live OpenClaw Benchmarks from Real Developer-Agent Sessions
- Attribution via Distributional Paths for Information Revelation
- Beyond Encoder Accumulation: Measuring Encoder Roles in Multi-Encoder VLMs
- Privacy-Robust Incrementality Measurement for Advertising Systems under Signal Loss
- MLP Splatting: Object-Centric Neural Fields
- From 'What' to 'How' and 'Why': Sharing LLM-Generated Retrospective Summaries of Older Adults' Passive Tracking Data with Remote Family Members
- Seg2Track++: Probabilistic Track Validation and Data Association for Multi-Object Tracking and Segmentation
- DyaPlex: Full-Duplex Speech-Motion Model for Dyadic Interaction
- Visual Instruction Tuning Aligns Modalities through Abstraction
- Taiji: Pareto Optimal Policy Optimization with Semantics-IDs Trade-off for Industrial LLM-Enhanced Recommendation
- Explainable Forecasting of Scientific Breakthroughs from Concept Network Dynamics
- FLARE: Fine-Grained Diagnostic Feedback for LLM Code Refinement
- Two-Action Apple Tasting with Switching Costs
- Clustered Self-Assessment: A Simple yet Effective Method for Uncertainty Quantification in Large Language Models
- Re-Evaluating Continual Learning with Few-Shot Adaptation
- Text-attributed Graph Condensation via Text Selection and Attribute Matching
- Where Do We (Not) Need Temporal Context in Low-Resource Video Task Adaptation?
- Online Learning with Gradient-Variation Interval Regret
- Dynamic Short Convolutions Improve Transformers
- Calibrating Urban Traffic Simulation from Sparse Road Observations via Genetic Optimization
- Finding Needles in the Haystack: Transductive Active Labeling in Ecology
- Leveraging BART to Assess CS1 C++ Programming Assignments using Rubric-based Criteria
- Consistency Training Can Entrench Misalignment
- TeX-1500: A Paired Real-World LWIR Hyperspectral Dataset and Benchmark for Temperature-Emissivity-Texture Decomposition
- LiveBand: Live Accompaniment Generation in the Audio Domain
- Template Collapse and Information-Theoretic Limits in Camera rPPG Pulse Morphology Restoration
- Signed Spiking Neuron Enabled by an Orthogonal-Easy-Axis Magnetic Tunnel Junction
- Limit Analysis of Graph Neural Networks with Wireless Conflict Graphs
- SLU-2K: A Question-Based Benchmark for Semantic Evaluation of Sign Language Translation
- AmbientEye: A Dataset for Pupil Segmentation under Natural Ambient Infrared Illumination
- KletterMix: Climbing Toward High-Quality German Pretraining Data - The Full Report
- Bregman meets L'evy: Stochastic mirror descent with heavy-tailed noise in continuous and discrete time
- Merit or networks? What decides where research is published
- Framing Migration News with LLMs: Structured CoT as a Support for Human Interpretation
- Neural Navigation Functions for Zero-Shot Generalizable Motion Planning
- Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models
- Qwen-Image-Flash: Rethinking the Training Recipe for Few-Step Distillation
- Predicting the Neutrino Mass Ordering Using Neural Networks
- Proof-Refactor: Refactoring Generated Formal Proofs into Modular Artifacts
- When to Re-Plan: Subgoal Persistence in Hierarchical Latent Reasoning
- Adaptive Inference for Resource-Constrained Dynamic Pricing
- Conformal Language Modeling via Posterior Sampling
- Compress then Merge: From Multiple LoRAs into One Low-Rank Adapter
- Unveiling the Structure of Do-Calculus Reasoning via Derivation Graphs
- Text-to-Image Models Need Less from Text Encoders Than You Think
- When Graph Tokens Sink: A Mechanistic Analysis of Graph Language Models
- Dynamic Objective Selection with Safeguards and LLM Oversight for Financial Decision-Making
- Don't Forget Your Embeddings: Robust Knowledge Erasure via Precise Editing of Embeddings
- Face versus Body Tracking for Human-Robot Interaction: An Egocentric Dataset
- SkillPyramid: A Hierarchical Skill Consolidation Framework for Self-Evolving Agents
- Staying Alive: Uncensored Survival Analysis with Tabular Foundation Models
- Speedrunning Tabular Foundation Model Pretraining
- A Fast Methane Detection Pipeline on Board Satellites Based on Mag1c-SAS and LinkNet
- Beyond Single Solution: Multi-Hypothesis Collaborative Deep Unfolding Network for Image Compressive Sensing
- AUGUSTE: Online-Learning dApp for Predictive URLLC Scheduling
- Towards Non-Monotonic Entailment in Propositional Defeasible Standpoint Logic
- Graph Regularized Non-negative Reduced Biquaternion Matrix Factorization for Color Image Recognition
- CoEval: Ranking Language Models for Custom Tasks Without Labeled Data or Trustworthy Benchmarks
- Safety Measurements for Fine-tuned LLMs Should be Grounded in Capability
- A Benchmark for Semi-supervised Multi-modal Crowd Counting
- Gender-Dependent Diagnostic Substitution in LLM Medical Triage: Same Symptoms, Unequal Urgency
- VidMsg: A Benchmark for Implicit Message Inference in Short Videos
- Building Reliable Long-Form Generation via Hallucination Rejection Sampling
- SkelHCC: A Hyperbolic CLIP-Driven Cache Adaptation Framework for Skeleton-based One-Shot Action Recognition
- CauTion: Knowing When to Trust LLMs for Ensemble Causal Discovery
- Set-Preserving Calibration from Conformal P-Values to E-Values
- Multimodal Transformer Based Generic Mixture Density Network for Scattering Timescale Estimation of Fast Radio Bursts
- Training a Predictive Coding Network on ImageNet using Equilibrium Propagation
- UnsOcc: 3D Semantic Occupancy Prediction in Unstructured Scene via Rendering Fusion
- AutoTail-BSFGM: Class-Balance-Aware Fine-Tuning for Chinese Scholarly Text Classification
- Few-Shot Prediction for Pulsar Noise with Long Short-Term Memory Network
- Learned Non-Maximum Suppression for 3D Object Detection
- Efficient Transformer-Based Localized Patch Sampling for Choroid Plexus Segmentation in Multiple Sclerosis
- Analytical Evaluation of DCA Convergence Properties for Minimizing Prediction Functions of Gaussian RBF Support Vector Regression
- From Prompt to Service: An SLM-Based Agent Orchestration Gateway for AI-Driven Virtual Worlds
- Constraint-Enhanced Physical Search through Correlation Matching
- A Robust Optimization Approach to Sparse Principal Component Analysis
- How Many Trees in a Random Forest? A Revisited Approach with Plateau Search and Optuna Integration
- SAGE: A Quantitative Evaluation of Socialized Evolution in Agent Ecosystems
- Attend to Anything: Foundation Model for Unified Human Attention Modeling
- Knowledge-Preserved Model Tuning in Null-Space for Robust Spatio-Temporal Video Grounding
- Can LLM Rerankers Predict Their Own Ranking Performance?
- High-Precision APT Malware Attribution with Out-of-Scope Resilience
- Scalable On-Hardware Training of Quantum Neural Networks and Application to Clinical Data Imputation
- Structure-Guided Mixed Masked Pretraining and Spatial Continuity Regularization for Printed Circuit Board Defect Detection
- AvatarMix: Identity-Preserving Cross-Avatar Composition for Outfit Personalization
- BaltiVoice: A Speech Corpus and Fine-tuned Whisper ASR System for the Balti Language
- Demystifying Pipeline Parallelism: First Theory for PipeDream
- HiSE: A Lightweight Hierarchical Semantic Explainer for Heterogeneous Graph Neural Networks
- Low-Frequency Shortcuts in Texture-Driven Visual Learning
- TrAction: Action Recognition with Sparse Trajectories
- Analyzing Stream Collapse in Hyper-Connections: From Diagnosis to Mitigation
- PersistGS: Differentiable Physics for Object Permanence in 4D Gaussian Splatting
- A formal definition and meta-model for a machine theory of mind
- Mixed-Modality Dual Face-Hair Retrieval
- Topology-Aware Gaussian Graph Repair for Robust Graph Neural Networks
- What Makes Interaction Trajectories Effective for Training Terminal Agents?
- From 3D Perception to Safety Reasoning: A Graph-Based Framework for Real-Time Underground Mine Monitoring
- Tonal parsimony in chord-sequence analysis: combining modulation cost and tonal vocabulary
- PerchRL: Vision-Based Agile Perching on Inclined Platforms under Rapid and Irregular Motion
- Large Language Models Are Overconfident in Their Own Responses
- A Hybrid Approach For Malware Classification Using Secondary Features Fusion
- FlowGuard: Flow Matching for Identity-Independent Detection of Data-Free Model Stealing Attacks on Energy System Intrusion Detection Systems
- PrimeSVT: An Automated Memory-aware Pruning Framework with Prioritized Compression Policy for Spiking Vision Transformers
- PHAF-Personalized Hand Avatars in a Flash
- Optimizing Explicit Unit-Distance Lower-Bound Certificates
- IDO: Incongruity-aware Distribution Optimization for Multimodal Fake News Detection
- A unified multi-task framework enables interpretable chest radiograph analysis
- Lexicons and grammars for language processing: industrial or handcrafted products?
- SAMatcher: Dense Co-Visibility Modeling via Cross-View Fusion for Scale-Imbalance Image Matching
- Towards Characterizing Scientific Image Utility and Upgradability
- Selective Token-Level Cryptographic Redaction for Privacy-Preserving Clinical Deployment of Large Language Models
- Causal Evidence of Stack Representations in Modeling Counter Languages Using Transformers
- Grasp-Then-Plan with Failure Attribution: A Closed Two-Stage Framework for Precise and Generalizable Robotic Manipulation
- AI Model Extraction Attacks: Bypassing Single-Client Assumptions in Defenses
- Link Prediction or Perdition: the Seeds of Instability in Knowledge Graph Embeddings
- EntSQL: A Benchmark for Grounding Text-to-SQL in Long-Context Enterprise Knowledge
- Speech Emotion Recognition using Attention-based LSTM-Network with Residual Connection
- The Impact of Temporal Granularity on Socio-Demographic Inference from Household Load Profiles
- The Unsampled Truth: Quantifying Prompt Artifacts in LM Psychometrics
- APIC: Amortized Physics-Informed Calibration using Neural Processes
- SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation
- Beyond Semantics: Modeling Factual and Affective Perceptual Experiences from Vision-Language Data
- RogueMerge: Robust and Unified Attacks against LLM Model Merging
- IdEst: Assessing Self-Supervised Learning Representations via Intrinsic Dimension
- Lingo_Research_Group at SemEval-2026 Task 9: Evaluating Prompt Variants for Polarization Detection
- Tailoring Strictly Proper Scoring Rules for Downstream Tasks: An Application to Causal Inference
- Evaluating LLMs' Effectiveness on Real-World Consumer Device Repair Questions
- FLIPS: Instance-Fingerprinting for LLMs via Pseudo-random Sequences
- Rethinking Calibration for High-Sparsity LLM Pruning: From Single Sources to Capability Coverage
- The Violation Situation Pattern: Persistent Representation of Compliance Violations in Knowledge Graphs
- Implement Kubernetes Pod-Level Remote Attestation for Confidential Workloads on dstack
- Multi-Modal Graph Neural Network with Transformer-Guided Adaptive Diffusion for Preclinical Alzheimer Classification
- Validation-Gated Multi-Agent Governance for Online Adaptation of Thermal-Hydraulic Surrogate Models under Operating-Regime Shift
- Beyond Ideal Instruction: A Comprehensive Framework for Evaluating LLMs in Realistic Interactions
- SPR: Toward a Graph Foundation Model for Transferable Graph Cognition via Spectral Patterns and Relational Geometry
- TASE: Truncation-Aware Semantic Embeddings for 3D Scene Understanding and Editing
- RobotValues: Evaluating Household Robots When Human Values Conflict
- Learning Multi-Scale Hypergraph for High-Order Brain Connectivity Analysis
- The Reliability Gap in Benchmark Auditing: Distribution Shift and Scale as Failure Modes of Contamination Detection
- From Script to Semantics: Prompting Strategies for African NLI
- Combining Statistical Features and Deep Encodings for Rehearsal-Based Class-Incremental Time Series Classification
- Message Tuning Outshines Graph Prompt Tuning: A Prismatic Space Perspective
- AI-Generated Traces for Novice Programmers: Learning Effects and Learner Differences in a Multi-Institutional Study
- BA-T: An Iterative Transformer for Two-View Bundle Adjustment
- SEA-NLI: Natural Language Inference as a Lens into Southeast Asian Cultural Understanding
- A Negative Result on Cross-Model Activation Transfer in a Pythia Multi-Hop Setting
- A Geometric Lens on Physics-Aligned Data Compression
- Are Common Substructures Transferable? Riemannian Graph Foundation Model with Neural Vector Bundles
- Distilling Answer-Set Programming Rules from LLMs for Neurosymbolic Visual Question Answering
- PaddleOCR-VL-1.6: Expanding the Frontier of Document Parsing with Under-Optimized Region Refinement and Progressive Post-Training
- Let There Be Light: Reflection, Refraction and Scattering for Neural Operators
- EqGINO: Equivariant Geometry-Informed Fourier Neural Operators for 3D PDEs
- Beyond "To whom it may concern": Tailoring Machine Translation to Audience and Intent
- PSViT: A Methodology for Structurally Pruning Spiking Vision Transformers
- OF3GS: On-the-Fly Feed-Forward 3D Gaussian Splatting from Unposed Images
- Do Real-World Datasets Contain Natural Experiments? An Empirical Study Using Causal Feature Selection
- The Word and the Way: Strategies for Domain-Specific BERT Pre-Training in German Medical NLP
- Structures Facilitate Retrieve, Rerank, and Generate
- MariData: One-Step Unpaired Image Translation for Maritime Environments
- Hierarchies of Calibration: Classification meets Regression
- When Does Complexity Conditioning Help a Frozen Sentence Embedding? A Controlled Study of Per-Sentence and Pair-Level Difficulty Adaptation
- MemoGen: Can Past Experience Improve Future Text-to-Image Generation?
- Benchmarking Speech-to-Speech Translation Models
- Solipsistic Superintelligence is Unlikely to be Cooperative
- GFFMERGE: Efficient Merging of Graph Neural Force Fields and Beyond
- Learning Temporal Causal Structure via Smooth Differentiable Optimization
- BotDirector: Robot Storytelling Across the Symmetrical Reality with Multi-modal Interactions
- WebRISE: Requirement-Induced State Evaluation for MLLM-Generated Web Artifacts
- Sample-Size Scaling of the African Languages NLI Evaluation
- Effect of Demographic Bias on Skin Lesion Classification
- Optimized Labeling Resource Allocation for Prediction-Assisted Inference via OPAL
- Critical evaluation of PINN for FWD inverse analysis and differentiable FEM as an alternative
- DECA: Decentralizing Block-Wise Adam for Efficient LLM Full-Parameter Fine-Tuning on Non-IID Data
- Fast Organic Crystal Structure Prediction with Unit Cell Flow Matching
- AI Rater Discrimination Depends on Scoring Protocol in Complex Clinical Decision-Making
- SenseJudge: Human-Centric Preference-Driven Judgment Framework
- FinStressTS: A Parametric Synthetic Benchmark for Time-Series Forecasting in Finance
- HyperPatch: Sequential Knowledge Editing Under n-ary Structural Drift
- Ask When It Pays: Cost-Aware Open-Ended Interaction for Instance Goal Navigation
- Auditing Engagement Incentives in the Kidfluencer Ecosystem: A Multimodal Weak Supervision Approach
- SketchSong: Hierarchical Song Generation with Sketch Planning and Fine-Grained Multi-Track Modeling
- JAVEDIT: Joint Audio-Visual Instruction-Guided Video Editing with Agentic Data Curation
- OpenAgenet / OAN Yellow Paper: Technical Architecture for Trust-Governed Resource Identity and Discovery
- OpenAgenet / OAN White Paper: Open Infrastructure for Trusted Agent Interconnection
- SRENet: Spectral Re-Entry Network for Point Cloud Action Recognition
- ClinicalMC: A Benchmark for Multi-Course Clinical Decision-Making with Large Language Models
- A cross-domain tropical species dataset with Chinese vernacular names and CITES source links
- A2: Smaller Self-Supervised ViTs Localize Better than Larger Ones
- How Visible Are Silent Manipulation Failures? An Observability Study of False-Success Detection in Simulated Robot Episodes
- DMT-CBT: Longitudinal Therapeutic State Modeling for CBT Counseling
- Synthetic Hallucinations, Real Gains: Hard Negatives from Frontier Models for FIM Hallucination Mitigation
- Decoupled Smart Contract Audits: Lightweight LLM Framework via Distillation and Aggregation
- Rethinking Neural Width for Alternating Current Optimal Power Flow Proxies
- TiWeaver: Unified Temporal Dynamics Modeling via Contextual Patching
- KC-3DGS: Kurtosis-Constrained Gaussian Splatting for High-Fidelity View Synthesis
- Learning to See via Epiretinal Implant Stimulation in silico with Model-Based Deep Reinforcement Learning
- AnyAudio-Judge: A Dynamic Rubric-Based Benchmark and Evaluator for Audio Instruction Following
- FAF-CD: Frequency-Aware Fusion for Change Detection under Imperfect Multimodal Remote Sensing
- Experience-Driven Dynamic Exits for LLMs with Reinforcement Learning
- Trans GAN-WT: A Feature Extraction and Interactive Learning-Based Anomaly Detection Model for Wind Turbine Time Series Data
- From Long News to Accurate Forecast: Importance-Aware Fusion and PRM-Guided Reflection for Time Series Forecasting
- Can Factual Opinions Be Edited (Manipulated) in Large Language Models?
- BAHSD: Bridging the Long-tail Gap via Adaptive Distillation in Black-box Sequential Recommendation
- Learning to Solve, Forgetting to Retain: Correct-Set Turnover in RLVR
- Multi-component Causal Tracing in Large Language Models
- Hierarchical Federated Learning with Dynamic Clustering and Adaptive Regularization for Robust Infrastructure Inspection
- Regret Pre-training: Bridging Prior and Posterior Views for Enhanced Knowledge Grounding
- G^2C-MT: Graph-Guided Context Selection for Document-Level Machine Translation
- Efficient Hyperparameter Optimization for LLM Reinforcement Learning
- ASymPO: Asymmetric-Scale Policy Optimization for Asynchronous LLM Post-Training Without Behavior Information
- ROBUST-WT: Robust Uncertainty-aware Segmentation Transform via Whitening and Training Enhancements
- Learn When and Where to Connect: Adaptive Virtual Nodes for Dynamic Message Passing on Graphs
- Trajectory-Aware Node Contributions and the Limits of Static Controllability
- ZX-Calculus:Trace-Indexed Dependent Types and Epistemic Semantics
- Brief Announcement: Generative Markov Model for Distributed Computing Systems
- FCUS-rPPG: A Fast-Converging Unsupervised Framework for Remote Photoplethysmography via Gradient Oscillation Suppression
- The Value Function Semi-Algebraic Set in Partially Observable Markov Decision Processes
- The Geometry of LLM-as-Judge: Why Inter-LLM Consensus Is Not Human Alignment
- RelGT-AC: A Relational Graph Transformer for Autocomplete Tasks in Relational Databases
- Will Accurate Fields Mislead Photonic Design? FromGlobal Accuracy to Port Readout
- TriEval: A Resource-Efficient Pipeline for LLM Bias, Toxicity, and Truthfulness Assessment
- Local and Global Contraction Principles for MCMC Mixing
- Conditional Hypothesis Generation for LLM-Based Text Analysis with Researcher-Specified Covariates
- SEA-LION-Embedding: Open and Reproducible Text Embeddings for Southeast Asia
- Spike-Aware INT8 Execution for Spiking Language Models on Commodity CPUs
- Hallucinations as Orthogonal Noise: Inference-Time Manifold Alignment via Dynamic Contextual Orthogonalization
- Reproducibility is the New Copyleft: Defining AGI-oriented Reproducible Builds
- A Fast Screening Approach for High-dimensional Outcomes and High-dimensional Predictors
- CoughSense: Five-Class Respiratory Disease Classification via Whisper Encoder Fine-Tuning and Dual-Encoder Cross-Attention Fusion with Balanced Contrastive Learning
- MARIO: Motion-Augmented Real-Time Multi-Sensor Inertial Odometry
- Neural Networks Provably Learn Spectral Representations for Group Composition
- A Language Model from 1913: Pretraining on Historical Text
- A Locally Deployed RAG-Based Academic Advising System for Course Selection
- DriftSched: Adaptive QoS-Aware Scheduling under Runtime Token Drift for Multi-Tenant GPU Inference
- Towards Compact Autonomous Driving Perception with Balanced Learning and Multi-sensor Fusion
- Memory Retrieval for Changing Preferences
- WISE-HAR: A Generalizable Ensemble Deep Learning Framework for WiFi-Based Human Activity Recognition
- Chatbots Output Meaningful (but Problematic) Language
- EURO-5K: When Does Domain Pretraining Matter? Benchmarking Transformers for EU Reporting Obligation Extraction
- Glass Box at Orbit: A Constitutional AI Verification Framework for Trustworthy Autonomous CubeSat Intelligence