Reinforcement Learning

Flaming-hot Initiation with Regular Execution Sampling for Large Language Models NAACL 2025 FlightGPT: Towards Generalizable and Interpretable UAV Vision-and-Language Navigation with Vision-Language Models EMNLP 2025 Focused-DPO: Enhancing Code Generation Through Focused Preference Optimization on Error-Prone Points ACL 2025 Formally Verified Approximate Policy Iteration AAAI 2025 Formal Quality Measures for Predictors in Markov Decision Processes AAAI 2025 Fourier Guided Adaptive Adversarial Augmentation for Generalization in Visual Reinforcement Learning AAAI 2025 FRACTAL: Fine-Grained Scoring from Aggregate Text Labels ACL 2025 From General Reward to Targeted Reward: Improving Open-ended Long-context Generation Models EMNLP 2025 From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment ACL 2025 Full-Step-DPO: Self-Supervised Preference Optimization with Step-wise Rewards for Mathematical Reasoning ACL 2025 GenAL: Generative Agent for Adaptive Learning AAAI 2025 Generation of Geodesics with Actor-Critic Reinforcement Learning to Predict Midpoints JMLR 2025 Generative Reward Modeling via Synthetic Criteria Preference Learning ACL 2025 GENMANIP: LLM-driven Simulation for Generalizable Instruction-Following Manipulation CVPR 2025 Google Translate’s Research Submission to WMT2025 EMNLP 2025 Governance in Motion: Co-evolution of Constitutions and AI models for Scalable Safety EMNLP 2025 Graph-Reward-SQL: Execution-Free Reinforcement Learning for Text-to-SQL via Graph Matching and Stepwise Reward EMNLP 2025 Group-Aware Reinforcement Learning for Output Diversity in Large Language Models EMNLP 2025 GROVE: A Generalized Reward for Learning Open-Vocabulary Physical Skill CVPR 2025 GRPO-LEAD: A Difficulty-Aware Reinforcement Learning Approach for Concise Mathematical Reasoning in Language Models EMNLP 2025 GRSN: Gated Recurrent Spiking Neurons for POMDPs and MARL AAAI 2025 GTR: Guided Thought Reinforcement Prevents Thought Collapse in RL-based VLM Agent Training ICCV 2025 HAF-RM: A Hybrid Alignment Framework for Reward Model Training ACL 2025 HAVER: Instance-Dependent Error Bounds for Maximum Mean Estimation and Applications to Q-Learning and Monte Carlo Tree Search AISTATS 2025 HelpSteer3: Human-Annotated Feedback and Edit Data to Empower Inference-Time Scaling in Open-Ended General-Domain Tasks ACL 2025