详细论文列表：

1. Oral

Spotlight

[1]. Generalized Policy Iteration using Tensor Approximation for Hybrid Control

[2]. A Theoretical Explanation of Deep RL Performance in Stochastic Environments

[3]. A Benchmark on Robust Semi-Supervised Learning in Open Environments

[4]. Generative Adversarial Inverse Multiagent Learning

[5]. AMAGO: Scalable In-Context Reinforcement Learning for Adaptive Agents

[6]. Confronting Reward Model Overoptimization with Constrained RLHF

[7]. Improved Efficiency Based on Learned Saccade and Continuous Scene Reconstruction From Foveated Visual Sampling

[8]. Harnessing Density Ratios for Online Reinforcement Learning

[9]. Proximal Policy Gradient Arborescence for Quality Diversity Reinforcement Learning

[10]. Social Reward: Evaluating and Enhancing Generative AI through Million-User Feedback from an Online Creative Community

[11]. Improving Offline RL by Blending Heuristics

[12]. Tool-Augmented Reward Modeling

[13]. Reward-Consistent Dynamics Models are Strongly Generalizable for Offline Reinforcement Learning

[14]. Towards Optimal Regret in Adversarial Linear MDPs with Bandit Feedback

[15]. Dual RL: Unification and New Methods for Reinforcement and Imitation Learning

[16]. Stabilizing Contrastive RL: Techniques for Robotic Goal Reaching from Offline Data

[17]. Safe RLHF: Safe Reinforcement Learning from Human Feedback

[19]. Blending Imitation and Reinforcement Learning for Robust Policy Improvement

[20]. On the Role of General Function Approximation in Offline Reinforcement Learning

[21]. Beyond Worst-case Attacks: Robust RL with Adaptive Defense via Non-dominated Policies

[22]. Massively Scalable Inverse Reinforcement Learning for Route Optimization

[23]. Bandits Meet Mechanism Design to Combat Clickbait in Online Recommendation

[24]. Towards Principled Representation Learning from Videos for Reinforcement Learning

[25]. TorchRL: A data-driven decision-making library for PyTorch

[26]. Towards Robust Offline Reinforcement Learning under Diverse Data Corruption

[27]. DyST: Towards Dynamic Neural Scene Representations on Real-World Videos

[28]. Impact of Computation in Integral Reinforcement Learning for Continuous-Time Control

[29]. Maximum Entropy Heterogeneous-Agent Reinforcement Learning

[30]. Learning Hierarchical World Models with Adaptive Temporal Abstractions from Discrete Latent Dynamics

[31]. Text2Reward: Dense Reward Generation with Language Models for Reinforcement Learning

[32]. Submodular Reinforcement Learning

[33]. Query-Policy Misalignment in Preference-Based Reinforcement Learning

[34]. Kernel Metric Learning for In-Sample Off-Policy Evaluation of Deterministic RL Policies

[35]. Provable Offline Preference-Based Reinforcement Learning

[36]. Provable Reward-Agnostic Preference-Based Reinforcement Learning

[37]. Entity-Centric Reinforcement Learning for Object Manipulation from Pixels

[38]. Constrained Bi-Level Optimization: Proximal Lagrangian Value function Approach and Hessian-free Algorithm

[39]. Addressing Signal Delay in Deep Reinforcement Learning

[40]. DrM: Mastering Visual Reinforcement Learning through Dormant Ratio Minimization

[41]. RealChat-1M: A Large-Scale Real-World LLM Conversation Dataset

[42]. EfficientDM: Efficient Quantization-Aware Fine-Tuning of Low-Bit Diffusion Models

[43]. SocioDojo: Building Lifelong Analytical Agents with Real-world Text and Time Series

[44]. Quasi-Monte Carlo for 3D Sliced Wasserstein

[45]. Cascading Reinforcement Learning

[46]. Task Adaptation from Skills: Information Geometry, Disentanglement, and New Objectives for Unsupervised Reinforcement Learning

[47]. Efficient Distributed Training with Full Communication-Computation Overlap

[48]. PTaRL: Prototype-based Tabular Representation Learning via Space Calibration

[50]. Physics-Regulated Deep Reinforcement Learning: Invariant Embeddings

[51]. Retroformer: Retrospective Large Language Agents with Policy Gradient Optimization

[52]. Open-ended VQA benchmarking of Vision-Language models by exploiting Classification datasets and their semantic hierarchy

[53]. ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs

[54]. SEAL: A Framework for Systematic Evaluation of Real-World Super-Resolution

[55]. BarLeRIa: An Efficient Tuning Framework for Referring Image Segmentation

[56]. Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Vision

[57]. TD-MPC2: Scalable, Robust World Models for Continuous Control

[58]. Adaptive Rational Activations to Boost Deep Reinforcement Learning

[59]. Robust Adversarial Reinforcement Learning via Bounded Rationality Curricula

Oral

[1]. Locality Sensitive Sparse Encoding for Learning World Models Online

[2]. Demonstration-Regularized RL

[3]. KoLA: Carefully Benchmarking World Knowledge of Large Language Models

[4]. On Representation Complexity of Model-based and Model-free Reinforcement Learning

[5]. RepoBench: Benchmarking Repository-Level Code Auto-Completion Systems

[6]. Policy Rehearsing: Training Generalizable Policies for Reinforcement Learning

[7]. NP-GL: Extending Power of Nature from Binary Problems to Real-World Graph Learning

[8]. Pessimistic Nonlinear Least-Squares Value Iteration for Offline Reinforcement Learning

[9]. Improving Language Models with Advantage-based Offline Policy Gradients

[10]. Dichotomy of Early and Late Phase Implicit Biases Can Provably Induce Grokking

[11]. PromptAgent: Strategic Planning with Language Models Enables Expert-level Prompt Optimization

[12]. Large Language Models as Automated Aligners for benchmarking Vision-Language Models

[13]. Reverse Diffusion Monte Carlo

[14]. PlaSma: Procedural Knowledge Models for Language-based Planning and Re-Planning

[15]. Towards Foundational Models for Molecular Learning on Large-Scale Multi-Task Datasets

[16]. Training Diffusion Models with Reinforcement Learning

[17]. Finite-Time Analysis of On-Policy Heterogeneous Federated Reinforcement Learning

[18]. Federated Q-Learning: Linear Regret Speedup with Low Communication Cost

[19]. The Trickle-down Impact of Reward Inconsistency on RLHF

[20]. Maximum Entropy Model Correction in Reinforcement Learning

[21]. Simple Hierarchical Planning with Diffusion

[22]. Regularized Robust MDPs and Risk-Sensitive MDPs: Equivalence, Policy Gradient, and Sample Complexity

[23]. Curriculum reinforcement learning for quantum architecture search under hardware errors

[24]. Variance-aware Regret Bounds for Stochastic Contextual Dueling Bandits

[25]. Directly Fine-Tuning Diffusion Models on Differentiable Rewards

[26]. Tree Search-Based Policy Optimization under Stochastic Execution Delay

[27]. Offline RL with Observation Histories: Analyzing and Improving Sample Complexity

[28]. Understanding Hidden Context in Preference Learning: Consequences for RLHF

[29]. Eureka: Human-Level Reward Design via Coding Large Language Models

[30]. Active Retrosynthetic Planning Aware of Route Quality

[31]. Fiber Monte Carlo

[32]. Retrieval-Guided Reinforcement Learning for Boolean Circuit Minimization

[33]. Provable Benefits of Multi-task RL under Non-Markovian Decision Making Processes

[34]. Follow-the-Perturbed-Leader for Adversarial Bandits: Heavy Tails, Robustness, and Privacy

[35]. ViLMA: A Zero-Shot Benchmark for Linguistic and Temporal Grounding in Video-Language Models

[36]. Score Models for Offline Goal-Conditioned Reinforcement Learning

[37]. A Policy Gradient Method for Confounded POMDPs

[38]. Achieving Fairness in Multi-Agent MDP Using Reinforcement Learning

[39]. Escape Sky-high Cost: Early-stopping Self-Consistency for Multi-step Reasoning

[40]. Customizable Combination of Parameter-Efficient Modules for Multi-Task Learning

[41]. Hindsight PRIORs for Reward Learning from Human Preferences

[42]. Reward Model Ensembles Help Mitigate Overoptimization

[43]. Feasibility-Guided Safe Offline Reinforcement Learning

[44]. Compositional Conservatism: A Transductive Approach in Offline Reinforcement Learning

[45]. Flow to Better: Offline Preference-based Reinforcement Learning via Preferred Trajectory Generation

[46]. PAE: Reinforcement Learning from External Knowledge for Efficient Exploration

[47]. Yet Another ICU Benchmark: A Flexible Multi-Center Framework for Clinical ML

[48]. Identifying Policy Gradient Subspaces

[49]. Contextual Bandits with Online Neural Regression

[50]. PARL: A Unified Framework for Policy Alignment in Reinforcement Learning

[51]. SafeDreamer: Safe Reinforcement Learning with World Models

[52]. MetaCoCo: A New Few-Shot Classification Benchmark with Spurious Correlation

[53]. GnnX-Bench: Unravelling the Utility of Perturbation-based GNN Explainers through In-depth Benchmarking

[54]. Confidence-aware Reward Optimization for Fine-tuning Text-to-Image Models

[55]. Provably Efficient Iterated CVaR Reinforcement Learning with Function Approximation and Human Feedback

[56]. Goodhart's Law in Reinforcement Learning

[57]. Score Regularized Policy Optimization through Diffusion Behavior

[58]. Making RL with Preference-based Feedback Efficient via Randomization

[59]. Adaptive Regret for Bandits Made Possible: Two Queries Suffice

[60]. Negatively Correlated Ensemble Reinforcement Learning for Online Diverse Game Level Generation

[61]. Achieving Sample and Computational Efficient Reinforcement Learning by Action Space Reduction via Grouping

[62]. Demystifying Linear MDPs and Novel Dynamics Aggregation Framework

[63]. PandaLM: An Automatic Evaluation Benchmark for LLM Instruction Tuning Optimization

[64]. Steve-Eye: Equipping LLM-based Embodied Agents with Visual Perception in Open Worlds

[65]. Consistency Models as a Rich and Efficient Policy Class for Reinforcement Learning

[66]. Contrastive Preference Learning: Learning from Human Feedback without Reinforcement Learning

[67]. Privileged Sensing Scaffolds Reinforcement Learning

[68]. Learning Planning Abstractions from Language

[69]. Tailoring Self-Rationalizers with Multi-Reward Distillation

[70]. Building Cooperative Embodied Agents Modularly with Large Language Models

[71]. A Black-box Approach for Non-stationary Multi-agent Reinforcement Learning

[72]. CrossLoco: Human Motion Driven Control of Legged Robots via Guided Unsupervised Reinforcement Learning

[73]. Let Models Speak Ciphers: Multiagent Debate through Embeddings

[74]. Learning interpretable control inputs and dynamics underlying animal locomotion

[75]. Does Progress On Object Recognition Benchmarks Improve Generalization on Crowdsourced, Global Data?

[76]. Jumanji: a Diverse Suite of Scalable Reinforcement Learning Environments in JAX

[77]. Searching for High-Value Molecules Using Reinforcement Learning and Transformers

[78]. Exploiting Causal Graph Priors with Posterior Sampling for Reinforcement Learning

[79]. Towards Diverse Behaviors: A Benchmark for Imitation Learning with Human Demonstrations

[80]. Privately Aligning Language Models with Reinforcement Learning

[81]. On the Expressivity of Objective-Specification Formalisms in Reinforcement Learning

[82]. SMath input errorMath input errorMath input errorMath input errorMath input errorMath input error$Math input error$AC: Energy-Based Reinforcement Learning with Stein Soft Actor Critic

[83]. Robust Model-Based Optimization for Challenging Fitness Landscapes

[84]. Replay across Experiments: A Natural Extension of Off-Policy RL

[85]. BEND: Benchmarking DNA Language Models on Biologically Meaningful Tasks

[86]. Piecewise Linear Parametrization of Policies: Towards Interpretable Deep Reinforcement Learning

[87]. Time-Efficient Reinforcement Learning with Stochastic Stateful Policies

[88]. Open the Black Box: Step-based Policy Updates for Temporally-Correlated Episodic Reinforcement Learning

[89]. Incentivized Truthful Communication for Federated Bandits

[90]. Diffusion Generative Flow Samplers: Improving learning signals through partial trajectory optimization

[91]. On Trajectory Augmentations for Off-Policy Evaluation

[92]. Understanding the Effects of RLHF on LLM Generalisation and Diversity

[93]. Beyond Stationarity: Convergence Analysis of Stochastic Softmax Policy Gradient Methods

[94]. Delphic Offline Reinforcement Learning under Nonidentifiable Hidden Confounding

[95]. Prioritized Soft Q-Decomposition for Lexicographic Reinforcement Learning

[96]. GlucoBench: Curated List of Continuous Glucose Monitoring Datasets with Prediction Benchmarks

[97]. Incentive-Aware Federated Learning with Training-Time Model Rewards

[98]. Early Neuron Alignment in Two-layer ReLU Networks with Small Initialization

[99]. Sample-Efficiency in Multi-Batch Reinforcement Learning: The Need for Dimension-Dependent Adaptivity

[100]. Off-Policy Primal-Dual Safe Reinforcement Learning

[101]. STARC: A General Framework For Quantifying Differences Between Reward Functions

[102]. GAIA: a benchmark for General AI Assistants

[103]. Unleashing the Power of Pre-trained Language Models for Offline Reinforcement Learning

[104]. Discovering Temporally-Aware Reinforcement Learning Algorithms

[105]. Revisiting Data Augmentation in Deep Reinforcement Learning

[106]. Reward-Free Curricula for Training Robust World Models

[107]. Provable and Practical: Efficient Exploration in Reinforcement Learning via Langevin Monte Carlo

[108]. CPPO: Continual Learning for Reinforcement Learning with Human Feedback

[109]. Game-Theoretic Robust Reinforcement Learning Handles Temporally-Coupled Perturbations

[110]. Bandits with Replenishable Knapsacks: the Best of both Worlds

[111]. A Study of Generalization in Offline Reinforcement Learning

[112]. Diverse Projection Ensembles for Distributional Reinforcement Learning

[113]. MIntRec2.0: A Large-scale Benchmark Dataset for Multimodal Intent Recognition and Out-of-scope Detection in Conversations

[114]. RLIF: Interactive Imitation Learning as Reinforcement Learning

[115]. Bongard-OpenWorld: Few-Shot Reasoning for Free-form Visual Concepts in the Real World

[116]. Uni-O4: Unifying Online and Offline Deep Reinforcement Learning with Multi-Step On-Policy Optimization

[117]. FFB: A Fair Fairness Benchmark for In-Processing Group Fairness Methods

[118]. EasyTPP: Towards Open Benchmarking Temporal Point Processes

[119]. Uni-O4: Unifying Online and Offline Deep Reinforcement Learning with Multi-Step On-Policy Optimization

[120]. FFB: A Fair Fairness Benchmark for In-Processing Group Fairness Methods

[121]. EasyTPP: Towards Open Benchmarking Temporal Point Processes

[122]. Combinatorial Bandits for Maximum Value Reward Function under Value-Index Feedback

[123]. Alice Benchmarks: Connecting Real World Object Re-Identification with the Synthetic

[124]. Video Language Planning

[125]. Transformers as Decision Makers: Provable In-Context Reinforcement Learning via Supervised Pretraining

[126]. Learning Over Molecular Conformer Ensembles: Datasets and Benchmarks

[127]. Free from Bellman Completeness: Trajectory Stitching via Model-based Return-conditioned Supervised Learning

[128]. Diffusion Models for Multi-Task Generative Modeling

[129]. Neural Active Learning Beyond Bandits

[130]. Revisiting Plasticity in Visual Reinforcement Learning: Data, Modules and Training Stages

[131]. Sample-efficient Learning of Infinite-horizon Average-reward MDPs with General Function Approximation

[132]. Offline Data Enhanced On-Policy Policy Gradient with Provable Guarantees

[133]. SALMON: Self-Alignment with Principle-Following Reward Models

[134]. Test-Time Adaptation with CLIP Reward for Zero-Shot Generalization in Vision-Language Models

[135]. SemiReward: A General Reward Model for Semi-supervised Learning

[136]. Horizon-Free Regret for Linear Markov Decision Processes

[137]. On Differentially Private Federated Linear Contextual Bandits

[138]. Neural Neighborhood Search for Multi-agent Path Finding

[139]. Understanding when Dynamics-Invariant Data Augmentations Benefit Model-free Reinforcement Learning Updates

[140]. Sample Efficient Myopic Exploration Through Multitask Reinforcement Learning with Diverse Tasks

[141]. The Update Equivalence Framework for Decision-Time Planning

[142]. Learning Reusable Dense Rewards for Multi-Stage Tasks

[143]. Time Fairness in Online Knapsack Problems

[144]. On the Hardness of Constrained Cooperative Multi-Agent Reinforcement Learning

[145]. RLCD: Reinforcement Learning from Contrastive Distillation for LM Alignment

[146]. Reasoning with Latent Diffusion in Offline Reinforcement Learning

[147]. Low Rank Matrix Completion via Robust Alternating Minimization in Nearly Linear Time

[148]. Belief-Enriched Pessimistic Q-Learning against Adversarial State Perturbations

[149]. SmartPlay : A Benchmark for LLMs as Intelligent Agents

[150]. SOHES: Self-supervised Open-world Hierarchical Entity Segmentation

[151]. Robust NAS benchmark under adversarial training: assessment, theory, and beyond

[152]. SCHEMA: State CHangEs MAtter for Procedure Planning in Instructional Videos

[153]. DNABERT-2: Efficient Foundation Model and Benchmark For Multi-Species Genomes

[154]. Reward Design for Justifiable Sequential Decision-Making

[155]. Fast Value Tracking for Deep Reinforcement Learning

[156]. MAMBA: an Effective World Model Approach for Meta-Reinforcement Learning

[157]. Tree-Planner: Efficient Close-loop Task Planning with Large Language Models

[158]. LOQA: Learning with Opponent Q-Learning Awareness

[159]. Intelligent Switching for Reset-Free RL

[160]. On the Limitations of Temperature Scaling for Distributions with Overlaps

[161]. True Knowledge Comes from Practice: Aligning Large Language Models with Embodied Environments via Reinforcement Learning

[162]. Skill Machines: Temporal Logic Skill Composition in Reinforcement Learning

[163]. Who to imitate: Imitating desired behavior from divserse multi-agent datasets

[164]. SweetDreamer: Aligning Geometric Priors in 2D diffusion for Consistent Text-to-3D

[165]. Uni-RLHF: Universal Platform and Benchmark Suite for Reinforcement Learning with Diverse Human Feedback

[166]. Vision-Language Models are Zero-Shot Reward Models for Reinforcement Learning

[167]. Learning Multi-Agent Communication from Graph Modeling Perspective

[168]. Efficient Multi-agent Reinforcement Learning by Planning

[169]. Sample-Efficient Multi-Agent RL: An Optimization Perspective

[170]. CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discovery

[171]. SRL: Scaling Distributed Reinforcement Learning to Over Ten Thousand Cores

[172]. Plan-Seq-Learn: Language Model Guided RL for Solving Long Horizon Robotics Tasks

[173]. Robust Model Based Reinforcement Learning Using Math input errorMath input errorMath input errorMath input errorMath input errorMath input error$Math input error$ Adaptive Control

[174]. Learning Unsupervised World Models for Autonomous Driving via Discrete Diffusion

[175]. Parameter-Efficient Multi-Task Model Fusion with Partial Linearizeation

[176]. Horizon-free Reinforcement Learning in Adversarial Linear Mixture MDPs

[177]. Multi-task Learning with 3D-Aware Regularization

[178]. DMBP: Diffusion model based predictor for robust offline reinforcement learning against state observation perturbations

[179]. Alignment as Reward-Guided Search

[180]. Multi-Task Reinforcement Learning with Mixture of Orthogonal Experts

[181]. Retro-fallback: retrosynthetic planning in an uncertain world

[182]. Duolando: Follower GPT with Off-Policy Reinforcement Learning for Dance Accompaniment

[183]. AdaMerging: Adaptive Model Merging for Multi-Task Learning

[184]. MetaTool Benchmark: Deciding Whether to Use Tools and Which to Use

[185]. AdjointDPM: Adjoint Sensitivity Method for Gradient Backpropagation of Diffusion Probabilistic Models

[186]. Integrating Planning and Deep Reinforcement Learning via Automatic Induction of Task Substructures

[187]. LoTa-Bench: Benchmarking Language-oriented Task Planners for Embodied Agents

[188]. Outliers with Opposing Signals Have an Outsized Effect on Neural Network Optimization

[189]. Threshold-Consistent Margin Loss for Open-World Deep Metric Learning

[190]. Rethinking Adversarial Policies: A Generalized Attack Formulation and Provable Defense in RL

[191]. Learning Multi-Agent Communication with Contrastive Learning

[192]. Closing the Gap between TD Learning and Supervised Learning - A Generalisation Point of View.

[193]. On Stationary Point Convergence of PPO-Clip

[194]. Provably Efficient CVaR RL in Low-rank MDPs

[195]. COPlanner: Plan to Roll Out Conservatively but to Explore Optimistically for Model-Based RL

[196]. Transport meets Variational Inference: Controlled Monte Carlo Diffusions

[197]. In-context Exploration-Exploitation for Reinforcement Learning

[198]. The All-Seeing Project: Towards Panoptic Visual Recognition and Understanding of the Open World

[199]. TASK PLANNING FOR VISUAL ROOM REARRANGEMENT UNDER PARTIAL OBSERVABILITY

[200]. Optimal Sample Complexity for Average Reward Markov Decision Processes

[201]. DreamSmooth: Improving Model-based Reinforcement Learning via Reward Smoothing

[202]. Meta Inverse Constrained Reinforcement Learning: Convergence Guarantee and Generalization Analysis

[203]. Cleanba: A Reproducible and Efficient Distributed Reinforcement Learning Platform

[204]. Combining Spatial and Temporal Abstraction in Planning for Better Generalization

[205]. Decision Transformer is a Robust Contender for Offline Reinforcement Learning

[206]. ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate

[207]. Bridging State and History Representations: Understanding Self-Predictive RL

[208]. InstructDET: Diversifying Referring Object Detection with Generalized Instructions

[209]. Deep Reinforcement Learning Guided Improvement Heuristic for Job Shop Scheduling

[210]. GRAPH-CONSTRAINED DIFFUSION FOR END-TO-END PATH PLANNING

[211]. Efficient Backdoor Attacks for Deep Neural Networks in Real-world Scenarios

[212]. VertiBench: Advancing Feature Distribution Diversity in Vertical Federated Learning Benchmarks

[213]. Grounding Multimodal Large Language Models to the World

[214]. VFLAIR: A Research Library and Benchmark for Vertical Federated Learning

[215]. Stylized Offline Reinforcement Learning: Extracting Diverse High-Quality Behaviors from Heterogeneous Datasets

[216]. Sample-Efficient Learning of POMDPs with Multiple Observations In Hindsight

[217]. Pre-training with Synthetic Data Helps Offline Reinforcement Learning

[218]. AgentVerse: Facilitating Multi-Agent Collaboration and Exploring Emergent Behaviors

[219]. Efficient Planning with Latent Diffusion

[220]. A Benchmark Study on Calibration

[221]. Attention-Guided Contrastive Role Representations for Multi-agent Reinforcement Learning

[222]. Query-Dependent Prompt Evaluation and Optimization with Offline Inverse RL

[223]. Quantifying the Sensitivity of Inverse Reinforcement Learning to Misspecification

[224]. Byzantine Robust Cooperative Multi-Agent Reinforcement Learning as a Bayesian Game

[225]. AutoVP: An Automated Visual Prompting Framework and Benchmark

[226]. AutoCast++: Enhancing World Event Prediction with Zero-shot Ranking-based Context Retrieval

[227]. REValueD: Regularised Ensemble Value-Decomposition for Factorisable Markov Decision Processes

[228]. Language Model Self-improvement by Reinforcement Learning Contemplation

[229]. Towards Offline Opponent Modeling with In-context Learning

[230]. Early Stopping Against Label Noise Without Validation Data

[231]. Langevin Monte Carlo for strongly log-concave distributions: Randomized midpoint revisited