Amanda Askell

reinforcement learning from human feedback (2) zero-shot learning (1) few-shot learning (1) transfer learning (1) language model evaluation (1) instruction following (1) model behavior (1) language model alignment (1) model alignment (1) autoregressive model (1) language model (1) reward model (1) foundation model (1) supervised fine-tuning (1) inverse scaling (1) model-written evaluation (1) behavior discovery (1) natural language supervision (1) image text pair (1) contrastive learning (1)