Stable Answers, Unfinished Reasoning: Why Self-Consensus Is Not a Safe Early-Exit Signal
This work asks whether any consensus rule is both safe and token-saving, and whether one can be selected once and reused, and whether the answer is often a placeholder the model had not settled on.
Yun-Xian Mo, Dong-Hao Zhao, He-Jia Geng
· 0 citations