Wenqiang Wei

reinforcement learning (2) preference learning (1) information extraction (1) language model alignment (1) reinforcement learning from human feedback (1) token classification (1) model alignment (1) adversarial attack (1) reward hacking (1) retrieval-augmented generation (1) task adaptation (1) knowledge graph construction (1) alignment tuning (1) llm alignment (1) adversarial preference learning (1) token-level classification (1) large language model (1) neural network (1)