CoRLHF: Reinforcement learning from human feedback with cooperative policy-reward optimization for LLMs
Qi Liu, Zhuoyang Song, Yuxin Liang, Zejian Xie, Songxin Zhang, Songxin Zhang, Jiaxing Zhang, Yanjie Li
Topics & Concepts
Reinforcement learningComputer sciencePreferenceCeiling (cloud)Artificial intelligencePolicy learningFeedback loopMachine learningOptimization problemOptimization algorithmReinforcementIterative learning controlQ-learningHuman-in-the-loopSet (abstract data type)Bridge (graph theory)Iterative methodMathematical optimizationActive learning (machine learning)Preference learningCeiling effectRecommender Systems and TechniquesTopic ModelingReinforcement Learning in Robotics