Saved in:
Bibliographic Details
Main Authors: Han, Dong, Ai, Zhehong, Cai, Pengxiang, Lu, Shanya, Chen, Jianpeng, Ye, Zihao, Sun, Shuzhou, Gao, Ben, Ge, Lingli, Wang, Weida, Zhou, Xiangxin, Liu, Xihui, Su, Mao, Ouyang, Wanli, Bai, Lei, Zhou, Dongzhan, Xu, Tao, Li, Yuqiang, Zhang, Shufei
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.08736
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914145507475456
author Han, Dong
Ai, Zhehong
Cai, Pengxiang
Lu, Shanya
Chen, Jianpeng
Ye, Zihao
Sun, Shuzhou
Gao, Ben
Ge, Lingli
Wang, Weida
Zhou, Xiangxin
Liu, Xihui
Su, Mao
Ouyang, Wanli
Bai, Lei
Zhou, Dongzhan
Xu, Tao
Li, Yuqiang
Zhang, Shufei
author_facet Han, Dong
Ai, Zhehong
Cai, Pengxiang
Lu, Shanya
Chen, Jianpeng
Ye, Zihao
Sun, Shuzhou
Gao, Ben
Ge, Lingli
Wang, Weida
Zhou, Xiangxin
Liu, Xihui
Su, Mao
Ouyang, Wanli
Bai, Lei
Zhou, Dongzhan
Xu, Tao
Li, Yuqiang
Zhang, Shufei
contents Bayesian optimization (BO) is a powerful tool for scientific discovery in chemistry, yet its efficiency is often hampered by the sparse experimental data and vast search space. Here, we introduce ChemBOMAS: a large language model (LLM)-enhanced multi-agent system that accelerates BO through synergistic data- and knowledge-driven strategies. Firstly, the data-driven strategy involves an 8B-scale LLM regressor fine-tuned on a mere 1% labeled samples for pseudo-data generation, robustly initializing the optimization process. Secondly, the knowledge-driven strategy employs a hybrid Retrieval-Augmented Generation approach to guide LLM in dividing the search space while mitigating LLM hallucinations. An Upper Confidence Bound algorithm then identifies high-potential subspaces within this established partition. Across the LLM-refined subspaces and supported by LLM-generated data, BO achieves the improvement of effectiveness and efficiency. Comprehensive evaluations across multiple scientific benchmarks demonstrate that ChemBOMAS set a new state-of-the-art, accelerating optimization efficiency by up to 5-fold compared to baseline methods.
format Preprint
id arxiv_https___arxiv_org_abs_2509_08736
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ChemBOMAS: Accelerated BO in Chemistry with LLM-Enhanced Multi-Agent System
Han, Dong
Ai, Zhehong
Cai, Pengxiang
Lu, Shanya
Chen, Jianpeng
Ye, Zihao
Sun, Shuzhou
Gao, Ben
Ge, Lingli
Wang, Weida
Zhou, Xiangxin
Liu, Xihui
Su, Mao
Ouyang, Wanli
Bai, Lei
Zhou, Dongzhan
Xu, Tao
Li, Yuqiang
Zhang, Shufei
Machine Learning
Bayesian optimization (BO) is a powerful tool for scientific discovery in chemistry, yet its efficiency is often hampered by the sparse experimental data and vast search space. Here, we introduce ChemBOMAS: a large language model (LLM)-enhanced multi-agent system that accelerates BO through synergistic data- and knowledge-driven strategies. Firstly, the data-driven strategy involves an 8B-scale LLM regressor fine-tuned on a mere 1% labeled samples for pseudo-data generation, robustly initializing the optimization process. Secondly, the knowledge-driven strategy employs a hybrid Retrieval-Augmented Generation approach to guide LLM in dividing the search space while mitigating LLM hallucinations. An Upper Confidence Bound algorithm then identifies high-potential subspaces within this established partition. Across the LLM-refined subspaces and supported by LLM-generated data, BO achieves the improvement of effectiveness and efficiency. Comprehensive evaluations across multiple scientific benchmarks demonstrate that ChemBOMAS set a new state-of-the-art, accelerating optimization efficiency by up to 5-fold compared to baseline methods.
title ChemBOMAS: Accelerated BO in Chemistry with LLM-Enhanced Multi-Agent System
topic Machine Learning
url https://arxiv.org/abs/2509.08736