CITE: Anytime-Valid Statistical Inference in LLM Self-Consistency

Hirofumi Ota, Naoto Iwase, Yuki Ichihara, Junpei Komiyama, Masaaki Imaizumi · 2026

Read CITE: Anytime-Valid Statistical Inference in LLM Self-Consistency by Hirofumi Ota, Naoto Iwase, Yuki Ichihara, Junpei Komiyama, Masaaki Imaizumi (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-05873.txt.

Description

Hirofumi Ota and four co-authors give language model self consistency a rigorous stopping rule. Models often sample many answers and aggregate them, which improves reasoning, but knowing when enough samples have been drawn is hard. The answer set is not known in advance, and data dependent stopping can silently break error guarantees that fixed sample rules enjoy.

Their algorithm, CITE, certifies a prespecified answer as the unique mode of the model's response distribution, valid at any stopping time. It controls false certification at a chosen level under arbitrary data driven stopping, with no prior knowledge of the answer categories. The authors prove matching lower bounds on stopping time and extend the method to confidence weighted voting. Simulations and model experiments confirm the error control in practice.

Catalog record

CREATORHirofumi Ota, Naoto Iwase, Yuki Ichihara, Junpei Komiyama, Masaaki Imaizumi
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE303 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.05873

Browse the full library · About the Tower of Babel