Tomasz Korbak

language model (4) reinforcement learning (3) kl divergence (2) distribution matching (2) large language model (2) policy gradient (2) language model fine-tuning (2) reinforcement learning from human feedback (2) bayesian inference (1) imitation learning (1) language model alignment (1) signaling games (1) conditional generation (1) transfer learning (1) generative model (1) preference alignment (1) distribution learning (1) inductive bia (1) reward maximization (1) self-supervised learning (1)