Toward Scientific Reasoning in LLMs: Training from Expert Discussions via Reinforcement Learning

Re-assign community