metaTextGrad: Automatically optimizing language model optimizers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xu, Guowei, Yuksekgonul, Mert, Guestrin, Carlos, Zou, James
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908606135271424
author Xu, Guowei
Yuksekgonul, Mert
Guestrin, Carlos
Zou, James
author_facet Xu, Guowei
Yuksekgonul, Mert
Guestrin, Carlos
Zou, James
contents Large language models (LLMs) are increasingly used in learning algorithms, evaluations, and optimization tasks. Recent studies have shown that using LLM-based optimizers to automatically optimize model prompts, demonstrations, predictions themselves, or other components can significantly enhance the performance of AI systems, as demonstrated by frameworks such as DSPy and TextGrad. However, optimizers built on language models themselves are usually designed by humans with manual design choices; optimizers themselves are not optimized. Moreover, these optimizers are general purpose by design, to be useful to a broad audience, and are not tailored for specific tasks. To address these challenges, we propose metaTextGrad, which focuses on designing a meta-optimizer to further enhance existing optimizers and align them to be good optimizers for a given task. Our approach consists of two key components: a meta prompt optimizer and a meta structure optimizer. The combination of these two significantly improves performance across multiple benchmarks, achieving an average absolute performance improvement of up to 6% compared to the best baseline.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18524
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle metaTextGrad: Automatically optimizing language model optimizers
Xu, Guowei
Yuksekgonul, Mert
Guestrin, Carlos
Zou, James
Computation and Language
Large language models (LLMs) are increasingly used in learning algorithms, evaluations, and optimization tasks. Recent studies have shown that using LLM-based optimizers to automatically optimize model prompts, demonstrations, predictions themselves, or other components can significantly enhance the performance of AI systems, as demonstrated by frameworks such as DSPy and TextGrad. However, optimizers built on language models themselves are usually designed by humans with manual design choices; optimizers themselves are not optimized. Moreover, these optimizers are general purpose by design, to be useful to a broad audience, and are not tailored for specific tasks. To address these challenges, we propose metaTextGrad, which focuses on designing a meta-optimizer to further enhance existing optimizers and align them to be good optimizers for a given task. Our approach consists of two key components: a meta prompt optimizer and a meta structure optimizer. The combination of these two significantly improves performance across multiple benchmarks, achieving an average absolute performance improvement of up to 6% compared to the best baseline.
title metaTextGrad: Automatically optimizing language model optimizers
topic Computation and Language
url https://arxiv.org/abs/2505.18524