Papers

New on arXiv

Sat 15 Nov 2025

221 papers

Cacheback: Speculative Decoding With Nothing But CacheEfficient LLM inference and serving → Speculative decoding
Learning Straight Flows: Variational Flow Matching for Efficient GenerationDiffusion models → Flow matching, time series and science
Multi-Value Alignment for LLMs via Value Decorrelation and ExtrapolationLLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment
Efficient Mathematical Reasoning Models via Dynamic Pruning and Knowledge DistillationLLM reasoning → Prompting and chain-of-thought reasoning +1
Concept-Based Interpretability for Toxicity DetectionLLM safety and security → Safety alignment, refusal and guardrails
ExplainableGuard: Interpretable Adversarial Defense for Large Language Models Using Chain-of-Thought ReasoningLLM reasoning → Reasoning for agents, planning and code
Don't Think of the White Bear: Ironic Negation in Transformer Models Under Cognitive LoadLLM reasoning → Reasoning benchmarks and evaluation
Reasoning Text-to-Video Retrieval via Digital Twin Video Representations and Large Language ModelsLLM reasoning → Multimodal reasoning
Fast Reasoning Segmentation for Images and VideosLLM reasoning → Multimodal reasoning
Constructing and Interpreting Digital Twin Representations for Visual Reasoning via Reinforcement LearningLLM reasoning → Multimodal reasoning
SAC-MoE: Reinforcement Learning with Mixture-of-Experts for Control of Hybrid Dynamical Systems with UncertaintyDeep reinforcement learning methods → RL theory and sample complexity
Dynamic Reward Scaling for Multivariate Time Series Anomaly Detection: A VAE-Enhanced Reinforcement Learning ApproachDeep reinforcement learning methods → Deep RL algorithms and representations
Reward and Guidance through Rubrics: Promoting Exploration to Improve Multi-Domain ReasoningLLM reasoning → Reinforcement learning for reasoning
SpaceVLM: Sub-Space Modeling of Negation in Vision-Language ModelsMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
Optimal Self-Consistency for Efficient Reasoning with Large Language ModelsLLM reasoning → Test-time compute and efficient thinking
UpBench: A Dynamically Evolving Real-World Labor-Market Agentic Benchmark Framework Built for Human-Centric AILLM agents and tool use → General LLM agent frameworks and benchmarks
LiDAR-GS++:Improving LiDAR Gaussian Reconstruction via Diffusion PriorsDiffusion models → 3D, point cloud and scene generation
Privacy-Preserving Prompt Injection Detection for LLMs Using Federated Learning and Embedding-Based NLP ClassificationLLM safety and security → Prompt injection and agent security
ProofWright: Towards Agentic Formal Verification of CUDALLM agents and tool use → Software engineering coding agents
Cmprsr: Abstractive Token-Level Question-Agnostic Prompt CompressorEfficient LLM inference and serving → Prompt and context compression
D3ToM: Decider-Guided Dynamic Token Merging for Accelerating Diffusion MLLMsMultimodal LLMs / vision-language models → Image editing and quality assessment +1
MoralReason: Generalizable Moral Decision Alignment For LLM Agents Using Reasoning-Level Reinforcement LearningLLM reasoning → Reasoning benchmarks and evaluation +1
CrossVid: A Comprehensive Benchmark for Evaluating Cross-Video Reasoning in Multimodal Large Language ModelsLLM reasoning → Multimodal reasoning +1
Reinforcement Learning for Chemical Ordering in Alloy NanoparticlesDeep reinforcement learning methods → RL for energy, networks and operations
Mobile-Agent-RAG: Driving Smart Multi-Agent Coordination with Contextual Knowledge Empowerment for Long-Horizon Mobile AutomationLLM agents and tool use → GUI and mobile agents
Beyond World Models: Rethinking Understanding in AI ModelsWorld models → World models for language agents
RulePilot: An LLM-Powered Agent for Security Rule GenerationLLM agents and tool use → Agent threats, guardrails and governance
Suppressing VLM Hallucinations with Spectral Representation FilteringMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
AlignTree: Efficient Defense Against LLM Jailbreak AttacksLLM safety and security → Jailbreaking
FaNe: Towards Fine-Grained Cross-Modal Contrast with False-Negative Reduction and Text-Conditioned Sparse AttentionMultimodal LLMs / vision-language models → Medical vision-language models
MME-RAG: Multi-Manager-Expert Retrieval-Augmented Generation for Fine-Grained Entity Recognition in Task-Oriented DialoguesRetrieval-augmented generation → Conversational and speech RAG
Mixture of States: Routing Token-Level Dynamics for Multimodal GenerationDiffusion models → Image generation, text-to-image and editing
OmniSparse: Training-Aware Fine-Grained Sparse Attention for Long-Video MLLMsEfficient LLM inference and serving → KV cache optimization
MMRINet: Efficient Mamba-Based Segmentation with Dual-Path Refinement for Low-Resource MRI AnalysisState-space sequence models (S4, Mamba) → Mamba for medical image analysis
Chemistry-Enhanced Diffusion-Based Framework for Small-to-Large Molecular Conformation GenerationDiffusion models → Flow matching, time series and science
Reinforcement Learning for Charging Optimization of Inhomogeneous Dicke Quantum BatteriesDeep reinforcement learning methods → Quantum reinforcement learning
TSGDiff: Rethinking Synthetic Time Series Generation from a Pure Graph PerspectiveDiffusion models → Flow matching, time series and science
Multi-Agent Collaborative Fuzzing with Continuous Reflection for Smart Contracts Vulnerability DetectionLLM agents and tool use → Agent threats, guardrails and governance
CriticSearch: Fine-Grained Credit Assignment for Search Agents via a Retrospective CriticLLM agents and tool use → Web, search, and research agents
Rethinking Deep Alignment Through The Lens Of Incomplete LearningLLM reasoning → Reasoning for agents, planning and code +2
FIA-Edit: Frequency-Interactive Attention for Efficient and High-Fidelity Inversion-Free Text-Guided Image EditingDiffusion models → Image generation, text-to-image and editing
AttackVLA: Benchmarking Adversarial and Backdoor Attacks on Vision-Language-Action ModelsVision-language-action models → Safety and adversarial attacks +1
Seeing is Believing: Rich-Context Hallucination Detection for MLLMs via Backward Visual GroundingMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Deep Reinforcement Learning for Automated Stock Trading: An Ensemble StrategyDeep reinforcement learning methods → Adversarial robustness and RL in finance
Continuous-time Discrete-space Diffusion Model for RecommendationDiffusion models → Diffusion recommenders
MetaGDPO: Alleviating Catastrophic Forgetting with Metacognitive Knowledge through Group Direct Preference OptimizationLLM reasoning → Test-time compute and efficient thinking +1
Freeze, Share, Shrink: Rethinking the Action Backbone in Diffusion PoliciesVision-language-action models → Efficient VLA inference and adaptation +1
Adaptive Begin-of-Video Tokens for Autoregressive Video Diffusion ModelsDiffusion models → Video generation
From Scaling to Structured Expressivity: Rethinking Transformers for CTR PredictionLLM reasoning → Test-time compute and efficient thinking
Learning to Hear by Seeing: It's Time for Vision Language Models to Understand Artistic Emotion from Sight and SoundMultimodal LLMs / vision-language models → Multimodal emotion understanding
Treatment Stitching with Schr"odinger Bridge for Enhancing Offline Reinforcement Learning in Adaptive Treatment StrategiesOffline RL and imitation learning from data → Safe offline RL
ProAV-DiT: A Projected Latent Diffusion Transformer for Efficient Synchronized Audio-Video GenerationDiffusion models → Audio, speech and music
TextBO: Bayesian Optimization in Language Space for Eval-Efficient Self-Improving AISynthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training
Intelligent Collaborative Optimization for Rubber Tyre Film Production Based on Multi-path Differentiated Clipping Proximal Policy OptimizationDeep reinforcement learning methods → RL for energy, networks and operations
PipeDiT: Accelerating Diffusion Transformers in Video Generation with Task Pipelining and Model DecouplingDiffusion models → Video generation
BackWeak: Backdooring Knowledge Distillation Simply with Weak Triggers and Fine-tuningKnowledge distillation → Federated, robust and private distillation
BudgetLeak: Membership Inference Attacks on RAG Systems via the Generation Budget Side ChannelRetrieval-augmented generation → Privacy and security of RAG +1
Preference Learning from Physics-Based Feedback: Tuning Language Models to Design BCC/B2 SuperalloysLLM alignment / RLHF / preference optimisation → Preference optimization for molecular design
Striking the Right Balance between Compute and Copy: Improving LLM Inferencing Under Speculative DecodingEfficient LLM inference and serving → KV cache optimization
GCAgent: Long-Video Understanding via Schematic and Narrative Episodic MemoryMultimodal LLMs / vision-language models → Video understanding with MLLMs
Null-Space Diffusion Distillation Unlocks Speed, Fidelity and Realism in Lensless ImagingDiffusion models → Inverse problems and image restoration
CURE: Cultural Understanding and Reasoning Evaluation - A Framework for "Thick" Culture Alignment Evaluation in LLMsLLM reasoning → Reasoning benchmarks and evaluation
Adaptive Diagnostic Reasoning Framework for Pathology with Multimodal Large Language ModelsLLM reasoning → Clinical and biomedical applications +1
Look as You Think: Unifying Reasoning and Visual Evidence Attribution for Verifiable Document RAG via Reinforcement LearningLLM reasoning → Multimodal reasoning +2
Selecting Fine-Tuning Examples by Quizzing VLMsDiffusion models → Image generation, text-to-image and editing
Critical or Compliant? The Double-Edged Sword of Reasoning in Chain-of-Thought ExplanationsLLM reasoning → Prompting and chain-of-thought reasoning
Improving Autoformalization Using Direct Dependency RetrievalRetrieval-augmented generation → Agentic RAG, agents and memory
From Classification to Cross-Modal Understanding: Leveraging Vision-Language Models for Fine-Grained Renal PathologyMultimodal LLMs / vision-language models → Medical vision-language models
A Reasoning Paradigm for Named Entity RecognitionLLM reasoning → Multimodal reasoning
Quantile Q-Learning: Revisiting Offline Extreme Q-Learning with Quantile RegressionOffline RL and imitation learning from data → Offline RL algorithms +1
Noisy MRI Reconstruction via MAP Estimation with an Implicit Deep-Denoiser PriorDiffusion models → Medical imaging and segmentation
Image-POSER: Reflective RL for Multi-Expert Image Generation and EditingDeep reinforcement learning methods → RL with diffusion and flow models
Studying Illustrations in Manuscripts: An Efficient Deep-Learning Approach
FUSE : Failure-aware Usage of Subagent Evidence for MultiModal Search and Recommendation
Detection of Cyberbullying in GIF using AI
An Improved Ensemble-Based Machine Learning Model with Feature Optimization for Early Diabetes Prediction
A Fractional Variational Approach to Spectral Filtering Using the Fourier Transform
Semantics Meet Signals: Dual Codebook Representationl Learning for Generative Recommendation
GateRA: Token-Aware Modulation for Parameter-Efficient Fine-Tuning
EgoCogNav: Cognition-aware Human Egocentric Navigation
A novel strategy for multi-resource load balancing in agent-based systems
Multimodal AI for Body Fat Estimation: Computer Vision and Anthropometry with DEXA Benchmarks
Human-aligned Quantification of Numerical Data
irace-evo: Automatic Algorithm Configuration Extended With LLM-Based Code Evolution
Application of Graph Based Vision Transformers Architectures for Accurate Temperature Prediction in Fiber Specklegram Sensors
Known Meets Unknown: Mitigating Overconfidence in Open Set Recognition
Can LLMs Create Legally Relevant Summaries and Analyses of Videos?
Calibrated Decomposition of Aleatoric and Epistemic Uncertainty in Deep Features for Inference-Time Adaptation
CEDL: Centre-Enhanced Discriminative Learning for Anomaly Detection
From Phonemes to Meaning: Evaluating Large Language Models on Tamil
Leveraging Quantum-Based Architectures for Robust Diagnostics
AGGRNet: Selective Feature Extraction and Aggregation for Enhanced Medical Image Classification
Quantum Optimization Algorithms
Learning to Trust: Bayesian Adaptation to Varying Suggester Reliability in Sequential Decision Making
BitSnap: Checkpoint Sparsification and Quantization in LLM Training
MTMed3D: A Multi-Task Transformer-Based Model for 3D Medical Imaging
Changes in Real Time: Online Scene Change Detection with Multi-View Fusion
Explainable AI-Generated Image Detection RewardBench
More Than Irrational: Modeling Belief-Biased Agents
VoiceCraft-X: Unifying Multilingual, Voice-Cloning Speech Synthesis and Speech Editing

All 221