Xueru Wen

Reinforcement Learning (1) Reasoning (1)

large language model (4) preference learning (2) reward model (2) chain-of-thought reasoning (1) weak supervision (1) reinforcement learning from human feedback (1) iterative refinement (1) model alignment (1) foundation model (1) hallucination mitigation (1) on-policy learning (1) parameter efficiency (1) reasoning ability (1) human preference (1) human annotation (1) reward signal (1) chinese language (1) preference dataset (1) human alignment (1) self-critic mechanism (1)