Taming Overconfidence in LLMs: Reward Calibration in RLHF.

Jixuan Leng, Chengsong Huang, Banghua Zhu, Jiaxin Huang

Browse the full ICLR paper archive.