reward modeling

CulFiT: A Fine-grained Cultural-aware LLM Training Paradigm via Multilingual Critique Data Synthesis ACL 2025 Dialogue Is Not Enough to Make a Communicative BabyLM (But Neither Is Developmentally Inspired Reinforcement Learning) EMNLP 2025 Diverging Preferences: When do Annotators Disagree and do Models Know? ICML 2025 DORM: Preference Data Weights Optimization for Reward Modeling in LLM Alignment EMNLP 2025 DPL: Diverse Preference Learning Without A Reference Model NAACL 2025 Dynamic Scaling of Unit Tests for Code Reward Modeling ACL 2025 Earlier Tokens Contribute More: Learning Direct Preference Optimization From Temporal Decay Perspective ICLR 2025 Efficient Safety Alignment of Large Language Models via Preference Re-ranking and Representation-based Reward Modeling ACL 2025 Embedding Domain Knowledge for Large Language Models via Reinforcement Learning from Augmented Generation EMNLP 2025 Enhancing RLHF with Human Gaze Modeling EMNLP 2025 From General Reward to Targeted Reward: Improving Open-ended Long-context Generation Models EMNLP 2025 Generative Reward Modeling via Synthetic Criteria Preference Learning ACL 2025 Generative Verifiers: Reward Modeling as Next-Token Prediction ICLR 2025 Graders Should Cheat: Privileged Information Enables Expert-Level Automated Evaluations EMNLP 2025 Hierarchical Reward Modeling for Fault Localization in Large Code Repositories EMNLP 2025 Hybrid Preferences: Learning to Route Instances for Human vs. AI Feedback ACL 2025 Improving Medical Large Vision-Language Models with Abnormal-Aware Feedback ACL 2025 Improving Reward Models with Synthetic Critiques NAACL 2025 In-Dataset Trajectory Return Regularization for Offline Preference-based Reinforcement Learning AAAI 2025 LEGEND: Leveraging Representation Engineering to Annotate Safety Margin for Preference Datasets AAAI 2025 Linear Contextual Bandits With Interference ICML 2025 Linguistic Generalizability of Test-Time Scaling in Mathematical Reasoning ACL 2025 LLMSR@XLLM25: A Language Model-Based Pipeline for Structured Reasoning Data Construction ACL 2025 Long-Term Ad Memorability: Understanding & Generating Memorable Ads WACV 2025 MagicID: Hybrid Preference Optimization for ID-Consistent and Dynamic-Preserved Video Customization ICCV 2025