The Importance of Directional Feedback for LLM-based Optimizers

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Nie, Allen, Cheng, Ching-An, Kolobov, Andrey, Swaminathan, Adith
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917699281485824
author Nie, Allen
Cheng, Ching-An
Kolobov, Andrey
Swaminathan, Adith
author_facet Nie, Allen
Cheng, Ching-An
Kolobov, Andrey
Swaminathan, Adith
contents We study the potential of using large language models (LLMs) as an interactive optimizer for solving maximization problems in a text space using natural language and numerical feedback. Inspired by the classical optimization literature, we classify the natural language feedback into directional and non-directional, where the former is a generalization of the first-order feedback to the natural language space. We find that LLMs are especially capable of optimization when they are provided with {directional feedback}. Based on this insight, we design a new LLM-based optimizer that synthesizes directional feedback from the historical optimization trace to achieve reliable improvement over iterations. Empirically, we show our LLM-based optimizer is more stable and efficient in solving optimization problems, from maximizing mathematical functions to optimizing prompts for writing poems, compared with existing techniques.
format Preprint
id arxiv_https___arxiv_org_abs_2405_16434
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle The Importance of Directional Feedback for LLM-based Optimizers
Nie, Allen
Cheng, Ching-An
Kolobov, Andrey
Swaminathan, Adith
Artificial Intelligence
Computation and Language
Neural and Evolutionary Computing
We study the potential of using large language models (LLMs) as an interactive optimizer for solving maximization problems in a text space using natural language and numerical feedback. Inspired by the classical optimization literature, we classify the natural language feedback into directional and non-directional, where the former is a generalization of the first-order feedback to the natural language space. We find that LLMs are especially capable of optimization when they are provided with {directional feedback}. Based on this insight, we design a new LLM-based optimizer that synthesizes directional feedback from the historical optimization trace to achieve reliable improvement over iterations. Empirically, we show our LLM-based optimizer is more stable and efficient in solving optimization problems, from maximizing mathematical functions to optimizing prompts for writing poems, compared with existing techniques.
title The Importance of Directional Feedback for LLM-based Optimizers
topic Artificial Intelligence
Computation and Language
Neural and Evolutionary Computing
url https://arxiv.org/abs/2405.16434