Enregistré dans:
Détails bibliographiques
Auteurs principaux: Kim, Jinseok, Cho, Sukmin, Jeong, Soyeong, Kim, Sangyeop, Cho, Sungzoon
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2509.02377
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912566799761408
author Kim, Jinseok
Cho, Sukmin
Jeong, Soyeong
Kim, Sangyeop
Cho, Sungzoon
author_facet Kim, Jinseok
Cho, Sukmin
Jeong, Soyeong
Kim, Sangyeop
Cho, Sungzoon
contents Query Expansion (QE) improves retrieval performance by enriching queries with related terms. Recently, Large Language Models (LLMs) have been used for QE, but existing methods face a trade-off: generating diverse terms boosts performance but increases computational cost. To address this challenge, we propose Candidate Token Query Expansion (CTQE), which extracts diverse and relevant terms from a single LLM decoding pass by leveraging unselected candidate tokens. These tokens, though not part of the final output, are conditioned on the full query and capture useful information. By aggregating them, CTQE achieves both relevance and diversity without extra inference, reducing overhead and latency. Experiments show that CTQE delivers strong retrieval performance with significantly lower cost, outperforming or comparable to more expensive methods. Code is available at: https://github.com/bluejeans8/CTQE
format Preprint
id arxiv_https___arxiv_org_abs_2509_02377
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Upcycling Candidate Tokens of Large Language Models for Query Expansion
Kim, Jinseok
Cho, Sukmin
Jeong, Soyeong
Kim, Sangyeop
Cho, Sungzoon
Information Retrieval
Query Expansion (QE) improves retrieval performance by enriching queries with related terms. Recently, Large Language Models (LLMs) have been used for QE, but existing methods face a trade-off: generating diverse terms boosts performance but increases computational cost. To address this challenge, we propose Candidate Token Query Expansion (CTQE), which extracts diverse and relevant terms from a single LLM decoding pass by leveraging unselected candidate tokens. These tokens, though not part of the final output, are conditioned on the full query and capture useful information. By aggregating them, CTQE achieves both relevance and diversity without extra inference, reducing overhead and latency. Experiments show that CTQE delivers strong retrieval performance with significantly lower cost, outperforming or comparable to more expensive methods. Code is available at: https://github.com/bluejeans8/CTQE
title Upcycling Candidate Tokens of Large Language Models for Query Expansion
topic Information Retrieval
url https://arxiv.org/abs/2509.02377