PETS: A Principled Framework Towards Optimal Trajectory Allocation for Efficient Test-Time Self-Consistency

Zhangyi Liu, Huaizhi Qu *, Xiaowei Yin *, He Sun, Yanjun Han, Tianlong Chen, Zhun Deng · 2026

Read PETS: A Principled Framework Towards Optimal Trajectory Allocation for Efficient Test-Time Self-Consistency by Zhangyi Liu, Huaizhi Qu *, Xiaowei Yin *, He Sun, Yanjun Han, Tianlong Chen, Zhun Deng (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-2602-16745.txt.

Description

Zhangyi Liu and co-authors introduce PETS, a principled frame for optimal trajectory allocation in test-time self-consistency. Test-time scaling lifts scores by aggregating stochastic reasoning paths, but sample-efficient allocation under a fixed budget stays an open challenge. Central to the approach is the self-consistency rate, defined as agreement with the infinite-budget majority vote, which grounds allocation in rigorous theory.

In the offline setting, where all questions are known in advance, allocation is linked to crowdsourcing theory by treating reasoning traces as workers, yielding guarantees and an efficient majority-voting allocation algorithm. In the online streaming setting, where questions arrive one by one, a new method adapts budgets to question difficulty while keeping strong guarantees and low compute. Model builders get a principled way to spend test-time samples wisely.

Catalog record

CREATORZhangyi Liu, Huaizhi Qu *, Xiaowei Yin *, He Sun, Yanjun Han, Tianlong Chen, Zhun Deng
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE107 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-16745

Browse the full library · About the Tower of Babel