Saved in:
Bibliographic Details
Main Authors: Xu, Jitao, Zhou, Hongyun, Shen, Lei, Zhu, Conghui, Huang, Jin, Duan, Yitao
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2501.04393
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913640707260416
author Xu, Jitao
Zhou, Hongyun
Shen, Lei
Zhu, Conghui
Huang, Jin
Duan, Yitao
author_facet Xu, Jitao
Zhou, Hongyun
Shen, Lei
Zhu, Conghui
Huang, Jin
Duan, Yitao
contents Large Language Models (LLMs) can benefit from useful experiences to improve their performance on specific tasks. However, finding helpful experiences for different LLMs is not obvious, since it is unclear what experiences suit specific LLMs. Previous studies intended to automatically find useful experiences using LLMs, while it is difficult to ensure the effectiveness of the obtained experience. In this paper, we propose Stochastic Experience Optimization (SEO), an iterative approach that finds optimized model-specific experience without modifying model parameters through experience update in natural language. In SEO, we propose a stochastic validation method to ensure the update direction of experience, avoiding unavailing updates. Experimental results on three tasks for three LLMs demonstrate that experiences optimized by SEO can achieve consistently improved performance. Further analysis indicates that SEO-optimized experience can generalize to out-of-distribution data, boosting the performance of LLMs on similar tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2501_04393
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SEO: Stochastic Experience Optimization for Large Language Models
Xu, Jitao
Zhou, Hongyun
Shen, Lei
Zhu, Conghui
Huang, Jin
Duan, Yitao
Computation and Language
Large Language Models (LLMs) can benefit from useful experiences to improve their performance on specific tasks. However, finding helpful experiences for different LLMs is not obvious, since it is unclear what experiences suit specific LLMs. Previous studies intended to automatically find useful experiences using LLMs, while it is difficult to ensure the effectiveness of the obtained experience. In this paper, we propose Stochastic Experience Optimization (SEO), an iterative approach that finds optimized model-specific experience without modifying model parameters through experience update in natural language. In SEO, we propose a stochastic validation method to ensure the update direction of experience, avoiding unavailing updates. Experimental results on three tasks for three LLMs demonstrate that experiences optimized by SEO can achieve consistently improved performance. Further analysis indicates that SEO-optimized experience can generalize to out-of-distribution data, boosting the performance of LLMs on similar tasks.
title SEO: Stochastic Experience Optimization for Large Language Models
topic Computation and Language
url https://arxiv.org/abs/2501.04393