AI RESEARCH
More Test-Time Compute Can Hurt: Overestimation Bias in LLM Beam Search
arXiv CS.AI
•
ArXi:2603.15377v1 Announce Type: cross Wider beam search should improve LLM reasoning, but when should you stop widening? Prior work on beam width selection has focused on inference efficiency \citep{qin2025dsbd, freitag2017beam}, without analyzing whether wider search can \emph{hurt} output quality. We present an analysis, grounded in Extreme Value Theory, that answers this question. Beam selection over noisy scorer outputs