Population-Evolve: a Parallel Sampling and Evolutionary Method for LLM Math Reasoning
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.
Code (0)
등록된 구현이 없습니다.
Tasks
Computational EfficiencySimilar Papers 제목 키워드 기반
The limits of weak selection and large population size in evolutionary game theory
Evolutionary game theory is a mathematical approach to studying how social behaviors evolve. In many recent works, evolutionary competition between strategies is modeled as a stochastic process in a finite population. In…
Evolution of Societies via Reinforcement Learning
The universe involves many independent co-learning agents as an ever-evolving part of our observed environment. Yet, in practice, Multi-Agent Reinforcement Learning (MARL) applications are typically constrained to small,…
Multi-agent Reinforcement Learningreinforcement-learningReinforcement LearningEvolving a Multi-Population Evolutionary-QAOA on Distributed QPUs
Our research combines an Evolutionary Algorithm (EA) with a Quantum Approximate Optimization Algorithm (QAOA) to update the ansatz parameters, in place of traditional gradient-based methods, and benchmark on the Max-Cut …
Phylogeny-informed fitness estimation
Phylogenies (ancestry trees) depict the evolutionary history of an evolving population. In evolutionary computing, a phylogeny can reveal how an evolutionary algorithm steers a population through a search space, illumina…
DiagnosticDiversityEvolutionary AlgorithmsA Multi-objective Evolutionary Algorithm Based on Bi-population with Uniform Sampling for Neural Architecture Search
Neural architecture search (NAS) automates neural network design, improving efficiency over manual approaches. However, efficiently discovering high-performance neural network architectures that simultaneously optimize m…
Neural Architecture Search