Yexin Li

large language model (2) multi-agent reinforcement learning (1) zero-shot learning (1) few-shot learning (1) in-context learning (1) ai safety (1) reinforcement learning from human feedback (1) value alignment (1) human value (1) sparse autoencoder (1) object-centric representation (1) decentralized learning (1) demonstration selection (1) causal value graph (1) value steering (1) structural generalization (1) causal inference (1) generative preference learning (1)