Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models.

Zixiang Chen, Yihe Deng, Huizhuo Yuan, Kaixuan Ji, Quanquan Gu

Browse the full ICML paper archive.