arXivShiyuan Feng, Huan-ang Gao, Haohan Chi, Hanlin Wu, Zhilong Zhang, Zheng Jiang, Bingxiang He, Wei-Ying Ma, Ya-Qin Zhang, Hao ZhouMon, Jul 6, 2026, 10:59 AM PDT
score 17.1
New method lets weak AI teach strong AI how to reason
Original: Weak-to-Strong Generalization via Direct On-Policy Distillation
Source: arxiv.org ↗
Writing ELI5 summary…