Runji Lin

reinforcement learning (3) reward model (2) mathematical reasoning (2) process reward model (2) offline reinforcement learning (1) monte carlo estimation (1) on-policy learning (1) sequence model (1) inference efficiency (1) expert routing (1) encoder-decoder architecture (1) text-based game (1) critique model (1) chain of thought (1) critic model (1) error identification (1) large language model (1) llm ensemble (1) real-time strategy (1) multi-agent system (1)