Learning planning-based reasoning by trajectoriescollection and process reward synthesizing 논문리뷰

https://arxiv.org/abs/2402.00658