KABB: Knowledge-Aware Bayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systems

Jusheng Zhang, Zimeng Huang, Yijia Fan, Ningyuan Liu, Mingyan Li, Zhuojie Yang, Jiawei Yao, Jian Wang, Keze Wang
Proceedings of the 42nd International Conference on Machine Learning, PMLR 267:74966-74996, 2025.

Abstract

As scaling large language models faces prohibitive costs, multi-agent systems emerge as a promising alternative, though challenged by static knowledge assumptions and coordination inefficiencies. We introduce Knowledge-Aware Bayesian Bandits (KABB), a novel framework that enhances multi-agent system coordination through semantic understanding and dynamic adaptation. The framework features three key innovations: a customized knowledge distance model for deep semantic understanding, a dual-adaptation mechanism for continuous expert optimization, and a knowledge-aware Thompson Sampling strategy for efficient expert selection. Extensive evaluation demonstrates KABB achieves an optimal cost-performance balance, maintaining high performance while keeping computational demands relatively low in multi-agent coordination.

Cite this Paper

BibTeX
@InProceedings{pmlr-v267-zhang25aa, title = {{KABB}: Knowledge-Aware {B}ayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systems}, author = {Zhang, Jusheng and Huang, Zimeng and Fan, Yijia and Liu, Ningyuan and Li, Mingyan and Yang, Zhuojie and Yao, Jiawei and Wang, Jian and Wang, Keze}, booktitle = {Proceedings of the 42nd International Conference on Machine Learning}, pages = {74966--74996}, year = {2025}, editor = {Singh, Aarti and Fazel, Maryam and Hsu, Daniel and Lacoste-Julien, Simon and Berkenkamp, Felix and Maharaj, Tegan and Wagstaff, Kiri and Zhu, Jerry}, volume = {267}, series = {Proceedings of Machine Learning Research}, month = {13--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v267/main/assets/zhang25aa/zhang25aa.pdf}, url = {https://proceedings.mlr.press/v267/zhang25aa.html}, abstract = {As scaling large language models faces prohibitive costs, multi-agent systems emerge as a promising alternative, though challenged by static knowledge assumptions and coordination inefficiencies. We introduce Knowledge-Aware Bayesian Bandits (KABB), a novel framework that enhances multi-agent system coordination through semantic understanding and dynamic adaptation. The framework features three key innovations: a customized knowledge distance model for deep semantic understanding, a dual-adaptation mechanism for continuous expert optimization, and a knowledge-aware Thompson Sampling strategy for efficient expert selection. Extensive evaluation demonstrates KABB achieves an optimal cost-performance balance, maintaining high performance while keeping computational demands relatively low in multi-agent coordination.} }
Endnote
%0 Conference Paper %T KABB: Knowledge-Aware Bayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systems %A Jusheng Zhang %A Zimeng Huang %A Yijia Fan %A Ningyuan Liu %A Mingyan Li %A Zhuojie Yang %A Jiawei Yao %A Jian Wang %A Keze Wang %B Proceedings of the 42nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Aarti Singh %E Maryam Fazel %E Daniel Hsu %E Simon Lacoste-Julien %E Felix Berkenkamp %E Tegan Maharaj %E Kiri Wagstaff %E Jerry Zhu %F pmlr-v267-zhang25aa %I PMLR %P 74966--74996 %U https://proceedings.mlr.press/v267/zhang25aa.html %V 267 %X As scaling large language models faces prohibitive costs, multi-agent systems emerge as a promising alternative, though challenged by static knowledge assumptions and coordination inefficiencies. We introduce Knowledge-Aware Bayesian Bandits (KABB), a novel framework that enhances multi-agent system coordination through semantic understanding and dynamic adaptation. The framework features three key innovations: a customized knowledge distance model for deep semantic understanding, a dual-adaptation mechanism for continuous expert optimization, and a knowledge-aware Thompson Sampling strategy for efficient expert selection. Extensive evaluation demonstrates KABB achieves an optimal cost-performance balance, maintaining high performance while keeping computational demands relatively low in multi-agent coordination.
APA
Zhang, J., Huang, Z., Fan, Y., Liu, N., Li, M., Yang, Z., Yao, J., Wang, J. & Wang, K.. (2025). KABB: Knowledge-Aware Bayesian Bandits for Dynamic Expert Coordination in Multi-Agent Systems. Proceedings of the 42nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 267:74966-74996 Available from https://proceedings.mlr.press/v267/zhang25aa.html.

Related Material