Yihong Dong

code generation (9) large language model (6) preference optimization (2) benchmark evaluation (2) reinforcement learning with verifiable reward (2) named entity recognition (1) program verification (1) prompt engineering (1) machine unlearning (1) policy learning (1) direct preference optimization (1) model alignment (1) source code (1) output distribution (1) data leakage (1) domain knowledge (1) software engineering (1) evaluation benchmark (1) domain-specific evaluation (1) few-shot learning (1)