Yu Jin Kim

code generation (2) large language model (2) policy optimization (1) transfer learning (1) knowledge distillation (1) data augmentation (1) in-context learning (1) question answering (1) preference optimization (1) model pretraining (1) model forgetting (1) computational efficiency (1) reward prediction (1) knowledge graph (1) context window (1) token prediction (1) domain-specific pretraining (1) next token prediction (1) commonsense reasoning (1) rejection sampling (1)