Feifan Song

large language model (6) preference alignment (3) reinforcement learning from human feedback (2) human preference (2) supervised fine-tuning (2) human alignment (2) preference learning (1) density estimation (1) few-shot learning (1) data augmentation (1) in-context learning (1) chain-of-thought reasoning (1) text generation (1) feature matching (1) object counting (1) feature extraction (1) importance sampling (1) knowledge distillation (1) token efficiency (1) continuous optimization (1)