Population-Evolve: a Parallel Sampling and Evolutionary Method for LLM Math Reasoning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Yanzhi, Duan, Yitong, Zhang, Zhaoxi, He, Jiyan, Zheng, Shuxin
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911331890757632
author Zhang, Yanzhi
Duan, Yitong
Zhang, Zhaoxi
He, Jiyan
Zheng, Shuxin
author_facet Zhang, Yanzhi
Duan, Yitong
Zhang, Zhaoxi
He, Jiyan
Zheng, Shuxin
contents Test-time scaling has emerged as a promising direction for enhancing the reasoning capabilities of Large Language Models in last few years. In this work, we propose Population-Evolve, a training-free method inspired by Genetic Algorithms to optimize LLM reasoning. Our approach maintains a dynamic population of candidate solutions for each problem via parallel reasoning. By incorporating an evolve prompt, the LLM self-evolves its population in all iterations. Upon convergence, the final answer is derived via majority voting. Furthermore, we establish a unification framework that interprets existing test-time scaling strategies through the lens of genetic algorithms. Empirical results demonstrate that Population-Evolve achieves superior accuracy with low performance variance and computational efficiency. Our findings highlight the potential of evolutionary strategies to unlock the reasoning power of LLMs during inference.
format Preprint
id arxiv_https___arxiv_org_abs_2512_19081
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Population-Evolve: a Parallel Sampling and Evolutionary Method for LLM Math Reasoning
Zhang, Yanzhi
Duan, Yitong
Zhang, Zhaoxi
He, Jiyan
Zheng, Shuxin
Artificial Intelligence
Test-time scaling has emerged as a promising direction for enhancing the reasoning capabilities of Large Language Models in last few years. In this work, we propose Population-Evolve, a training-free method inspired by Genetic Algorithms to optimize LLM reasoning. Our approach maintains a dynamic population of candidate solutions for each problem via parallel reasoning. By incorporating an evolve prompt, the LLM self-evolves its population in all iterations. Upon convergence, the final answer is derived via majority voting. Furthermore, we establish a unification framework that interprets existing test-time scaling strategies through the lens of genetic algorithms. Empirical results demonstrate that Population-Evolve achieves superior accuracy with low performance variance and computational efficiency. Our findings highlight the potential of evolutionary strategies to unlock the reasoning power of LLMs during inference.
title Population-Evolve: a Parallel Sampling and Evolutionary Method for LLM Math Reasoning
topic Artificial Intelligence
url https://arxiv.org/abs/2512.19081