Shujin Wu

reinforcement learning (3) supervised fine-tuning (2) reasoning chain (2) hallucination mitigation (2) confidence calibration (2) large language model (2) multimodal large language model (1) multimodal reasoning (1) knowledge boundary (1) human-ai interaction (1) multi-turn conversation (1) contrastive preference optimization (1) proactive engagement (1) self-reflective rationale (1) conditional reinforcement learning (1) language model trustworthiness (1) self-rewarding signal (1) contrastive learning (1) self-reflective reasoning (1) uncertainty quantification (1)