Papers

New on arXiv

Wed 29 May 2024

383 papers

Conveyor: Efficient Tool-aware LLM Serving with Tool Partial ExecutionEfficient LLM inference and serving → Serving systems and request scheduling
Toward Conversational Agents with Context and Time Sensitive Long-term MemoryRetrieval-augmented generation → Conversational and speech RAG +1
Blind Image Restoration via Fast Diffusion InversionDiffusion models → Inverse problems and image restoration
Dr-LLaVA: Visual Instruction Tuning with Symbolic Clinical GroundingMultimodal LLMs / vision-language models → Medical vision-language models +1
Unlearning Climate Misinformation in Large Language ModelsLLM safety and security → Machine unlearning
RLeXplore: Accelerating Research in Intrinsically-Motivated Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
CLIPLoss and Norm-Based Data Selection Methods for Multimodal Contrastive LearningMultimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP
One-Shot Safety Alignment for Large Language Models via Optimal DualizationLLM alignment / RLHF / preference optimisation → Safety alignment and jailbreak attacks +1
Preference Learning Algorithms Do Not Learn Preference RankingsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Predicting Long-Term Human Behaviors in Discrete Representations via Physics-Guided DiffusionDiffusion models → Diffusion and flow policies for robotics
Two-Layer Retrieval-Augmented Generation Framework for Low-Resource Medical Question Answering Using Reddit Data: Proof-of-Concept StudyRetrieval-augmented generation → Medical and biomedical RAG
Momentum for the Win: Collaborative Federated Reinforcement Learning across Heterogeneous EnvironmentsDeep reinforcement learning methods → RL for energy, networks and operations
MemControl: Mitigating Memorization in Diffusion Models via Automated Parameter SelectionDiffusion models → Diffusion foundations and sampling
SSFF: Investigating LLM Predictive Capabilities for Startup Success through a Multi-Agent Framework with Enhanced Explainability and PerformanceLLM agents and tool use → Financial trading and forecasting agents
FourierMamba: Fourier Learning Integration with State Space Models for Image DerainingState-space sequence models (S4, Mamba) → Mamba for image restoration
MathChat: Benchmarking Mathematical Reasoning and Instruction Following in Multi-Turn InteractionsLLM reasoning → Mathematical reasoning
Adaptive In-conversation Team Building for Language Model AgentsLLM agents and tool use → General LLM agent frameworks and benchmarks
Diffusion Policy Attacker: Crafting Adversarial Attacks for Diffusion-based PoliciesDiffusion models → Robustness, privacy and adversarial aspects
Evaluating Vision-Language Models on Bistable ImagesMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Safety through Permissibility: Shield Construction for Fast and Safe Reinforcement LearningDeep reinforcement learning methods → Safe and constrained RL
X-VILA: Cross-Modality Alignment for Large Language ModelLLM alignment / RLHF / preference optimisation → Diffusion and multimodal model alignment
Self-Exploring Language Models: Active Preference Elicitation for Online AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Normative Modules: A Generative Agent Architecture for Learning Norms that Supports Multi-Agent CooperationLLM agents and tool use → Social simulation and multi-agent debate
Reasoning3D -- Grounding and Reasoning in 3D: Fine-Grained Zero-Shot Open-Vocabulary 3D Reasoning Part Segmentation via Large Vision-Language ModelsLLM reasoning → Spatial and embodied reasoning +1
Nearest Neighbor Speculative Decoding for LLM Generation and AttributionEfficient LLM inference and serving → Speculative decoding
Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHFLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Robust Preference Optimization through Reward Model DistillationLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Matryoshka Query Transformer for Large Vision-Language ModelsMultimodal LLMs / vision-language models → Visual token pruning and efficiency
Data Efficient Behavior Cloning for Fine Manipulation via Continuity-based Corrective LabelsOffline RL and imitation learning from data → Robot manipulation imitation learning
Adaptive Image Quality Assessment via Teaching Large Multimodal Model to CompareMultimodal LLMs / vision-language models → Image editing and quality assessment
Rich-Observation Reinforcement Learning with Continuous Latent DynamicsDeep reinforcement learning methods → Deep RL algorithms and representations
Weak-to-Strong Search: Align Large Language Models via Searching over Small Language ModelsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Faster Cascades via Speculative DecodingEfficient LLM inference and serving → Speculative decoding
ConceptPrune: Concept Editing in Diffusion Models via Skilled Neuron PruningDiffusion models → Robustness, privacy and adversarial aspects +1
Forward-Backward Knowledge Distillation for Continual ClusteringKnowledge distillation → Distillation for continual learning
Gone but Not Forgotten: Improved Benchmarks for Machine UnlearningLLM safety and security → Machine unlearning
VideoTree: Adaptive Tree-based Video Representation for LLM Reasoning on Long VideosLLM reasoning → Multimodal reasoning
A Multi-Source Retrieval Question Answering Framework Based on RAGRetrieval-augmented generation → RAG robustness and adaptive retrieval
Going beyond Compositions, DDPMs Can Produce Zero-Shot InterpolationsDiffusion models → Diffusion foundations and sampling
DyDiff: Long-Horizon Rollout via Dynamics Diffusion for Offline Reinforcement LearningDiffusion models → Diffusion and flow policies for robotics +1
MetaToken: Detecting Hallucination in Image Descriptions by Meta ClassificationMultimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Conditional Latent ODEs for Motion Prediction in Autonomous DrivingEnd-to-end autonomous driving → Imitation-learning end-to-end planners +1
Learning from Litigation: Graphs and LLMs for Retrieval and Reasoning in eDiscoveryLLM reasoning → Reasoning for agents, planning and code
A Study of Plasticity Loss in On-Policy Deep Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
Can Graph Learning Improve Planning in LLM-based Agents?LLM agents and tool use → General LLM agent frameworks and benchmarks
PathReasoner: Modeling Reasoning Path with Equivalent Extension for Logical Question AnsweringLLM reasoning → Reasoning benchmarks and evaluation
Offline Regularised Reinforcement Learning for Large Language Models AlignmentLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Voice Jailbreak Attacks Against GPT-4oLLM safety and security → Audio jailbreaks and speech backdoors
Cracking the Code of Juxtaposition: Can AI Models Understand the Humorous ContradictionsMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
MEMoE: Enhancing Model Editing with Mixture of Experts AdaptorsMixture-of-experts → MoE large language models
OMPO: A Unified Framework for RL under Policy and Dynamics ShiftsDeep reinforcement learning methods → Deep RL algorithms and representations
Cephalo: Multi-Modal Vision-Language Models for Bio-Inspired Materials Analysis and DesignMultimodal LLMs / vision-language models → Chart understanding and anomaly detection
Statistical Context Detection for Deep Lifelong Reinforcement LearningDeep reinforcement learning methods → Deep RL algorithms and representations
BLSP-KD: Bootstrapping Language-Speech Pre-training via Knowledge DistillationSynthetic data and self-improvement for LLMs → Self-rewarding alignment and self-training +1
DiveR-CT: Diversity-enhanced Red Teaming Large Language Model Assistants with Relaxing ConstraintsLLM safety and security → Prompt injection and agent security
Inverse Concave-Utility Reinforcement Learning is Inverse Game TheoryOffline RL and imitation learning from data → Inverse reinforcement learning +1
Efficient Exploration in Average-Reward Constrained Reinforcement Learning: Achieving Near-Optimal Regret With Posterior SamplingDeep reinforcement learning methods → RL theory and sample complexity
Trust the Model Where It Trusts Itself -- Model-Based Actor-Critic with Uncertainty-Aware Rollout AdaptionDeep reinforcement learning methods → RL theory and sample complexity
Enhancing Vision-Language Model with Unmasked Token AlignmentMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
EasyAnimate: High-Performance Video Generation Framework with Hybrid Windows Attention and Reward BackpropagationDiffusion models → Video generation +1
Robust Optimization in Protein Fitness Landscapes Using Reinforcement Learning in Latent SpaceDeep reinforcement learning methods → RL for energy, networks and operations
Optimizing Vehicular Networks with Variational Quantum Circuits-based Reinforcement LearningDeep reinforcement learning methods → Quantum reinforcement learning
Are You Sure? Rank Them Again: Repeated Ranking For Better Preference DatasetsLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Kestrel: 3D Multimodal LLM for Part-Aware Grounded DescriptionMultimodal LLMs / vision-language models → Embodied and 3D vision-language models
Towards Better Chain-of-Thought: A Reflection on Effectiveness and FaithfulnessLLM reasoning → Prompting and chain-of-thought reasoning
DecomCAM: Advancing Beyond Saliency Maps through Decomposition and IntegrationMultimodal LLMs / vision-language models → VLM benchmarks and visual reasoning
Inference-Time Alignment of Diffusion Models with Direct Noise OptimizationDiffusion models → Diffusion and flow policies for robotics
Enhancing Descriptive Image Quality Assessment with A Large-scale Multi-modal DatasetMultimodal LLMs / vision-language models → Image editing and quality assessment
MoNDE: Mixture of Near-Data Experts for Large-Scale Sparse ModelsMixture-of-experts → Efficient MoE training and inference systems
Flow Priors for Linear Inverse Problems via Iterative Corrupted Trajectory MatchingDiffusion models → Flow matching, time series and science
Tilde: Teleoperation for Dexterous In-Hand Manipulation Learning with a DeltaHandOffline RL and imitation learning from data → Robot manipulation imitation learning
Federated Q-Learning with Reference-Advantage Decomposition: Almost Optimal Regret and Logarithmic Communication CostDeep reinforcement learning methods → RL for energy, networks and operations
Policy Zooming: Adaptive Discretization-based Infinite-Horizon Average-Reward Reinforcement LearningDeep reinforcement learning methods → RL theory and sample complexity
Kernel Metric Learning for In-Sample Off-Policy Evaluation of Deterministic RL PoliciesDeep reinforcement learning methods → RL theory and sample complexity
Principled Probabilistic Imaging using Diffusion Models as Plug-and-Play PriorsDiffusion models → Inverse problems and image restoration
Certifying Counterfactual Bias in LLMsLLM safety and security → Safety alignment, refusal and guardrails
Multimodal Adversarial Defense for Vision-Language Models by Leveraging One-To-Many RelationshipsMultimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness
FDQN: A Flexible Deep Q-Network Framework for Game AutomationDeep reinforcement learning methods → Deep RL algorithms and representations
Multi-objective Cross-task Learning via Goal-conditioned GPT-based Decision Transformers for Surgical Robot Task AutomationOffline RL and imitation learning from data → Safe offline RL
T2V-Turbo: Breaking the Quality Bottleneck of Video Consistency Model with Mixed Reward FeedbackDiffusion models → Video generation +1
Reverse Image Retrieval Cues Parametric Memory in Multimodal LLMsMultimodal LLMs / vision-language models → Document understanding and multimodal RAG
Preferred-Action-Optimized Diffusion Policies for Offline Reinforcement LearningDiffusion models → Diffusion and flow policies for robotics +1
CtrlA: Adaptive Retrieval-Augmented Generation via Inherent ControlRetrieval-augmented generation → RAG robustness and adaptive retrieval
Efficient Model-agnostic Alignment via Bayesian PersuasionLLM alignment / RLHF / preference optimisation → LLM preference optimization methods
Calibrating Reasoning in Language Models with Internal ConsistencyLLM reasoning → Reasoning benchmarks and evaluation
Multi-Condition Latent Diffusion Network for Scene-Aware Neural Human Motion PredictionDiffusion models → Video generation
Spectral-Risk Safe Reinforcement Learning with Convergence GuaranteesDeep reinforcement learning methods → RL theory and sample complexity
Data-Efficient Approach to Humanoid Control via Fine-Tuning a Pre-Trained GPT on Action DataOffline RL and imitation learning from data → Real-world robot offline RL
Efficient Preference-based Reinforcement Learning via Aligned Experience EstimationLLM alignment / RLHF / preference optimisation → Reward modeling and RLHF +1
Advancing Household Robotics: Deep Interactive Reinforcement Learning for Efficient Training and Enhanced PerformanceDeep reinforcement learning methods → RL for robot locomotion and manipulation
Vim-F: Visual State Space Model Benefiting from Learning in the Frequency DomainState-space sequence models (S4, Mamba) → Vision Mamba backbones
Fast Explanations via Policy Gradient-Optimized ExplainerDeep reinforcement learning methods → Interpretable and logic-guided RL
Improved Robustness and Hyperparameter Selection in the Dense Associative Memory
Prototype Analysis in Hopfield Networks with Hebbian Learning
cryoSPHERE: Single-particle heterogeneous reconstruction from cryo EM
A Machine Learning Approach for Identifying Anatomical Biomarkers of Early Mild Cognitive Impairment
Rotation Averaging: A Primal-Dual Method and Closed-Forms in Cycle Graphs
Interdisciplinary Expertise to Advance Equitable Explainable AI
Prompting or Fine-tuning? Exploring Large Language Models for Causal Graph Validation
Sisteme Hibride de Invatare Automata si Aplicatii

All 383