作者位次 #6/6NeurIPS 2026Poster· Deep Learning / Generative Models and Autoencoders
You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectories
Zhepei Wei · Xinyu Zhu · Wei-Lin Chen · Chengsong Huang · Jiaxin Huang · Yu Meng
强化学习与后训练OpenReview