Adaptive Acquisition Selection for Bayesian Optimization with Large Language Models

2026-02-08

AI Summary

This work addresses the absence of a universally optimal acquisition function in Bayesian optimization and the limited ability of existing adaptive methods to leverage optimization state information for dynamic selection. To overcome this, we propose LMABO, a novel framework that, for the first time, employs a pre-trained large language model as a zero-shot online policy selector. By utilizing structured state prompts—encoding key aspects such as budget consumption and surrogate model status—the method dynamically chooses the most suitable acquisition function at each iteration. Integrating Bayesian optimization, multi-acquisition-function ensembles, and prompt engineering, LMABO demonstrates significant performance gains over static, adaptive, and existing large language model–based baselines across 50 benchmark problems, confirming its effectiveness and strong generalization capability.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Bayesian Optimization critically depends on the choice of acquisition function, but no single strategy is universally optimal; the best choice is non-stationary and problem-dependent. Existing adaptive portfolio methods often base their decisions on past function values while ignoring richer information like remaining budget or surrogate model characteristics. To address this, we introduce LMABO, a novel framework that casts a pre-trained Large Language Model (LLM) as a zero-shot, online strategist for the BO process. At each iteration, LMABO uses a structured state representation to prompt the LLM to select the most suitable acquisition function from a diverse portfolio. In an evaluation across 50 benchmark problems, LMABO demonstrates a significant performance improvement over strong static, adaptive portfolio, and other LLM-based baselines. We show that the LLM's behavior is a comprehensive strategy that adapts to real-time progress, proving its advantage stems from its ability to process and synthesize the complete optimization state into an effective, adaptive policy.

Problem

• Bayesian Optimization
• acquisition function selection
• adaptive strategy
• Large Language Models
• optimization state

Innovation

• Large Language Models
• Bayesian Optimization
• Adaptive Acquisition
• Zero-shot Strategy
• Optimization State Representation