@inproceedings{sun-etal-2026-large-language,
title = "Large Language Model-Enhanced Multi-Armed Bandits",
author = "Sun, Jiahang and
Wang, Zhiyong and
Yang, Runhan and
Xiao, Chenjun and
Lui, John C.s. and
Dai, Zhongxiang",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.368/",
pages = "8130--8145",
ISBN = "979-8-89176-390-6",
abstract = "Large language models (LLMs) have been applied to sequential decision-making tasks like multi-armed bandits (MAB), where an LLM is tasked with selecting arms in each iteration. However, this direct arm selection approach is often suboptimal. We propose an alternative method combining classical MAB algorithms with LLMs. Specifically, we use a classical MAB framework and leverage the in-context learning capability of LLMs for reward prediction. First, we integrate the LLM-based predictor into Thompson sampling (TS) with a decaying temperature schedule to balance exploration and exploitation. We also incorporate the predictor into a regression oracle-based MAB algorithm with explicit exploration. Additionally, we extend our TS-based algorithm to dueling bandits, where only preference feedback between arm pairs is available, requiring significant algorithmic modifications. Our empirical evaluations on synthetic MAB tasks show that our algorithms outperform LLM-based direct arm selection. In experiments on real-world text datasets, we demonstrate that, in tasks where arms lack exploitable semantic meaning, our approach delivers significantly better performance than direct arm selection."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sun-etal-2026-large-language">
<titleInfo>
<title>Large Language Model-Enhanced Multi-Armed Bandits</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jiahang</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhiyong</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Runhan</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chenjun</namePart>
<namePart type="family">Xiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">John</namePart>
<namePart type="given">C.s.</namePart>
<namePart type="family">Lui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhongxiang</namePart>
<namePart type="family">Dai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Large language models (LLMs) have been applied to sequential decision-making tasks like multi-armed bandits (MAB), where an LLM is tasked with selecting arms in each iteration. However, this direct arm selection approach is often suboptimal. We propose an alternative method combining classical MAB algorithms with LLMs. Specifically, we use a classical MAB framework and leverage the in-context learning capability of LLMs for reward prediction. First, we integrate the LLM-based predictor into Thompson sampling (TS) with a decaying temperature schedule to balance exploration and exploitation. We also incorporate the predictor into a regression oracle-based MAB algorithm with explicit exploration. Additionally, we extend our TS-based algorithm to dueling bandits, where only preference feedback between arm pairs is available, requiring significant algorithmic modifications. Our empirical evaluations on synthetic MAB tasks show that our algorithms outperform LLM-based direct arm selection. In experiments on real-world text datasets, we demonstrate that, in tasks where arms lack exploitable semantic meaning, our approach delivers significantly better performance than direct arm selection.</abstract>
<identifier type="citekey">sun-etal-2026-large-language</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.368/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>8130</start>
<end>8145</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Large Language Model-Enhanced Multi-Armed Bandits
%A Sun, Jiahang
%A Wang, Zhiyong
%A Yang, Runhan
%A Xiao, Chenjun
%A Lui, John C.s.
%A Dai, Zhongxiang
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F sun-etal-2026-large-language
%X Large language models (LLMs) have been applied to sequential decision-making tasks like multi-armed bandits (MAB), where an LLM is tasked with selecting arms in each iteration. However, this direct arm selection approach is often suboptimal. We propose an alternative method combining classical MAB algorithms with LLMs. Specifically, we use a classical MAB framework and leverage the in-context learning capability of LLMs for reward prediction. First, we integrate the LLM-based predictor into Thompson sampling (TS) with a decaying temperature schedule to balance exploration and exploitation. We also incorporate the predictor into a regression oracle-based MAB algorithm with explicit exploration. Additionally, we extend our TS-based algorithm to dueling bandits, where only preference feedback between arm pairs is available, requiring significant algorithmic modifications. Our empirical evaluations on synthetic MAB tasks show that our algorithms outperform LLM-based direct arm selection. In experiments on real-world text datasets, we demonstrate that, in tasks where arms lack exploitable semantic meaning, our approach delivers significantly better performance than direct arm selection.
%U https://aclanthology.org/2026.acl-long.368/
%P 8130-8145
Markdown (Informal)
[Large Language Model-Enhanced Multi-Armed Bandits](https://aclanthology.org/2026.acl-long.368/) (Sun et al., ACL 2026)
ACL
- Jiahang Sun, Zhiyong Wang, Runhan Yang, Chenjun Xiao, John C.s. Lui, and Zhongxiang Dai. 2026. Large Language Model-Enhanced Multi-Armed Bandits. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 8130–8145, San Diego, California, United States. Association for Computational Linguistics.