Xuekai Zhu

knowledge distillation (2) reinforcement learning (1) preference learning (1) preference optimization (1) text style transfer (1) loss landscape (1) instruction tuning (1) reinforcement learning from human feedback (1) synthetic datum (1) reward model (1) biomedical domain (1) supervised fine-tuning (1) medical domain (1) content preservation (1) reasoning chain (1) medical benchmark (1) style embedding (1) medical ai (1) large language model (1) advantage function (1)