RLHF Workflow: From Reward Modeling to Online RLHF
Paper
• 2405.07863 • Published
• 71
Reinforcement Learning (RL / RLHF)
-
-
Understanding and Diagnosing Deep Reinforcement Learning
Paper • 2406.16979 • Published • 10 -
Direct Nash Optimization: Teaching Language Models to Self-Improve with General Preferences
Paper • 2404.03715 • Published • 62 -
Iterative Nash Policy Optimization: Aligning LLMs with General Preferences via No-Regret Learning
Paper • 2407.00617 • Published • 7 -
Offline Regularised Reinforcement Learning for Large Language Models Alignment
Paper • 2405.19107 • Published • 15 -
DogeRM: Equipping Reward Models with Domain Knowledge through Model Merging
Paper • 2407.01470 • Published • 7 -
Understanding the performance gap between online and offline alignment algorithms
Paper • 2405.08448 • Published • 18 -
Value-Incentivized Preference Optimization: A Unified Approach to Online and Offline RLHF
Paper • 2405.19320 • Published • 10 -
OpenRLHF: An Easy-to-use, Scalable and High-performance RLHF Framework
Paper • 2405.11143 • Published • 41 -
Dataset Reset Policy Optimization for RLHF
Paper • 2404.08495 • Published • 9 -
WPO: Enhancing RLHF with Weighted Preference Optimization
Paper • 2406.11827 • Published • 17 -
LLaRA: Supercharging Robot Learning Data for Vision-Language Policy
Paper • 2406.20095 • Published • 18 -
DigiRL: Training In-The-Wild Device-Control Agents with Autonomous Reinforcement Learning
Paper • 2406.11896 • Published • 20 -
Measuring memorization in RLHF for code completion
Paper • 2406.11715 • Published • 7 -
Artificial Generational Intelligence: Cultural Accumulation in Reinforcement Learning
Paper • 2406.00392 • Published • 14 -
Gradient Boosting Reinforcement Learning
Paper • 2407.08250 • Published • 13 -
PERL: Parameter Efficient Reinforcement Learning from Human Feedback
Paper • 2403.10704 • Published • 60 -
Mixtures of Experts Unlock Parameter Scaling for Deep RL
Paper • 2402.08609 • Published • 36 -
RM-Bench: Benchmarking Reward Models of Language Models with Subtlety and Style
Paper • 2410.16184 • Published • 25