@inproceedings{hong-etal-2026-gr2,
title = "$\text{GR}^2$: Mitigating Granularity Collapse via Adaptive Query Routing in Generative Retrieval",
author = "Hong, Soona and
Lee, Dohyeon and
Kim, Jaeyoung and
Hwang, Seung-Won",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1928/",
pages = "43891--43902",
ISBN = "979-8-89176-459-0",
abstract = "Generative Retrieval (GR) aims to directly generate document identifiers (DocIDs) without external indices. We identify granularity collapse: during training, GR models develop a preference for generating DocIDs with a specific granularity, causing retrieval failures when it differs from that of the target document{'}s DocID. We propose to quantify this mismatch using ``isotropy'' and confirm that retrieval accuracy degrades as such granularity mismatch increases. To address this, we propose Granularity Routing for Generative Retrieval ($\text{GR}^2$), which trains multiple expert models specialized for different granularity levels and routes each query to the appropriate expert. Experiments on NQ320k demonstrate that improves retrieval performance, with larger gains on queries that suffer more from granularity collapse."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hong-etal-2026-gr2">
<titleInfo>
<title>\textGR²: Mitigating Granularity Collapse via Adaptive Query Routing in Generative Retrieval</title>
</titleInfo>
<name type="personal">
<namePart type="given">Soona</namePart>
<namePart type="family">Hong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dohyeon</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jaeyoung</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seung-Won</namePart>
<namePart type="family">Hwang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Generative Retrieval (GR) aims to directly generate document identifiers (DocIDs) without external indices. We identify granularity collapse: during training, GR models develop a preference for generating DocIDs with a specific granularity, causing retrieval failures when it differs from that of the target document’s DocID. We propose to quantify this mismatch using “isotropy” and confirm that retrieval accuracy degrades as such granularity mismatch increases. To address this, we propose Granularity Routing for Generative Retrieval (\textGR²), which trains multiple expert models specialized for different granularity levels and routes each query to the appropriate expert. Experiments on NQ320k demonstrate that improves retrieval performance, with larger gains on queries that suffer more from granularity collapse.</abstract>
<identifier type="citekey">hong-etal-2026-gr2</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1928/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>43891</start>
<end>43902</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T \textGR²: Mitigating Granularity Collapse via Adaptive Query Routing in Generative Retrieval
%A Hong, Soona
%A Lee, Dohyeon
%A Kim, Jaeyoung
%A Hwang, Seung-Won
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F hong-etal-2026-gr2
%X Generative Retrieval (GR) aims to directly generate document identifiers (DocIDs) without external indices. We identify granularity collapse: during training, GR models develop a preference for generating DocIDs with a specific granularity, causing retrieval failures when it differs from that of the target document’s DocID. We propose to quantify this mismatch using “isotropy” and confirm that retrieval accuracy degrades as such granularity mismatch increases. To address this, we propose Granularity Routing for Generative Retrieval (\textGR²), which trains multiple expert models specialized for different granularity levels and routes each query to the appropriate expert. Experiments on NQ320k demonstrate that improves retrieval performance, with larger gains on queries that suffer more from granularity collapse.
%U https://aclanthology.org/2026.emnlp-main.1928/
%P 43891-43902
Markdown (Informal)
[GR2: Mitigating Granularity Collapse via Adaptive Query Routing in Generative Retrieval](https://aclanthology.org/2026.emnlp-main.1928/) (Hong et al., EMNLP 2026)
ACL