Large Language Models As Evolution Strategies

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lange, Robert Tjarko, Tian, Yingtao, Tang, Yujin
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909122816901120
author Lange, Robert Tjarko
Tian, Yingtao
Tang, Yujin
author_facet Lange, Robert Tjarko
Tian, Yingtao
Tang, Yujin
contents Large Transformer models are capable of implementing a plethora of so-called in-context learning algorithms. These include gradient descent, classification, sequence completion, transformation, and improvement. In this work, we investigate whether large language models (LLMs), which never explicitly encountered the task of black-box optimization, are in principle capable of implementing evolutionary optimization algorithms. While previous works have solely focused on language-based task specification, we move forward and focus on the zero-shot application of LLMs to black-box optimization. We introduce a novel prompting strategy, consisting of least-to-most sorting of discretized population members and querying the LLM to propose an improvement to the mean statistic, i.e. perform a type of black-box recombination operation. Empirically, we find that our setup allows the user to obtain an LLM-based evolution strategy, which we call `EvoLLM', that robustly outperforms baseline algorithms such as random search and Gaussian Hill Climbing on synthetic BBOB functions as well as small neuroevolution tasks. Hence, LLMs can act as `plug-in' in-context recombination operators. We provide several comparative studies of the LLM's model size, prompt strategy, and context construction. Finally, we show that one can flexibly improve EvoLLM's performance by providing teacher algorithm information via instruction fine-tuning on previously collected teacher optimization trajectories.
format Preprint
id arxiv_https___arxiv_org_abs_2402_18381
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Large Language Models As Evolution Strategies
Lange, Robert Tjarko
Tian, Yingtao
Tang, Yujin
Artificial Intelligence
Machine Learning
Neural and Evolutionary Computing
Large Transformer models are capable of implementing a plethora of so-called in-context learning algorithms. These include gradient descent, classification, sequence completion, transformation, and improvement. In this work, we investigate whether large language models (LLMs), which never explicitly encountered the task of black-box optimization, are in principle capable of implementing evolutionary optimization algorithms. While previous works have solely focused on language-based task specification, we move forward and focus on the zero-shot application of LLMs to black-box optimization. We introduce a novel prompting strategy, consisting of least-to-most sorting of discretized population members and querying the LLM to propose an improvement to the mean statistic, i.e. perform a type of black-box recombination operation. Empirically, we find that our setup allows the user to obtain an LLM-based evolution strategy, which we call `EvoLLM', that robustly outperforms baseline algorithms such as random search and Gaussian Hill Climbing on synthetic BBOB functions as well as small neuroevolution tasks. Hence, LLMs can act as `plug-in' in-context recombination operators. We provide several comparative studies of the LLM's model size, prompt strategy, and context construction. Finally, we show that one can flexibly improve EvoLLM's performance by providing teacher algorithm information via instruction fine-tuning on previously collected teacher optimization trajectories.
title Large Language Models As Evolution Strategies
topic Artificial Intelligence
Machine Learning
Neural and Evolutionary Computing
url https://arxiv.org/abs/2402.18381