@inproceedings{jiang-etal-2026-vibe,
title = "{VIBE}-Bench: Evaluating Personalized Large Language Models When Profiles Don{'}t Mean Preferences",
author = "Jiang, Yiwen and
Deng, Yang and
Fong, Stephanie and
Wang, Zimu and
Shen, Yaling and
Feng, Wei and
Yang, Hongxi and
Zhao, Xiangyu and
Xu, Zhongxing and
Mehta, Deval and
Cheng, Xuelian and
Ge, Zongyuan",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1676/",
pages = "35771--35797",
ISBN = "979-8-89176-462-0",
abstract = "Personalized Large Language Models (PLLMs) aim to tailor responses to individual users, where a central challenge is preference reasoning: inferring query-relevant preferences from user-related history. Existing benchmarks, however, largely assume that such preference can be retrieved from semantically related history. We study an underexplored but practically important regime, profile-preference conceptual misalignment (PRCM), where observable profile cues and query-specific preferences lie in different concept spaces, making semantic retrieval inconsistent for personalization. We introduce Vibe-Bench, a benchmark with two psychology-grounded tasks, 3,504 personas and 12,239 dialogues, including a manually verified gold test set, and requires cross-concept preference reasoning beyond surface semantic overlap. Experiments with several personalization methods show that current PLLMs largely rely on shallow semantic correlations and fail to acquire robust cross-concept mappings. These findings establish PRCM as a distinct failure regime in PLLMs and position Vibe-Bench as a focused testbed for advancing preference reasoning beyond semantic matching."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jiang-etal-2026-vibe">
<titleInfo>
<title>VIBE-Bench: Evaluating Personalized Large Language Models When Profiles Don’t Mean Preferences</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yiwen</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yang</namePart>
<namePart type="family">Deng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Stephanie</namePart>
<namePart type="family">Fong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zimu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yaling</namePart>
<namePart type="family">Shen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Feng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hongxi</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangyu</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhongxing</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Deval</namePart>
<namePart type="family">Mehta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xuelian</namePart>
<namePart type="family">Cheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zongyuan</namePart>
<namePart type="family">Ge</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Personalized Large Language Models (PLLMs) aim to tailor responses to individual users, where a central challenge is preference reasoning: inferring query-relevant preferences from user-related history. Existing benchmarks, however, largely assume that such preference can be retrieved from semantically related history. We study an underexplored but practically important regime, profile-preference conceptual misalignment (PRCM), where observable profile cues and query-specific preferences lie in different concept spaces, making semantic retrieval inconsistent for personalization. We introduce Vibe-Bench, a benchmark with two psychology-grounded tasks, 3,504 personas and 12,239 dialogues, including a manually verified gold test set, and requires cross-concept preference reasoning beyond surface semantic overlap. Experiments with several personalization methods show that current PLLMs largely rely on shallow semantic correlations and fail to acquire robust cross-concept mappings. These findings establish PRCM as a distinct failure regime in PLLMs and position Vibe-Bench as a focused testbed for advancing preference reasoning beyond semantic matching.</abstract>
<identifier type="citekey">jiang-etal-2026-vibe</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1676/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>35771</start>
<end>35797</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T VIBE-Bench: Evaluating Personalized Large Language Models When Profiles Don’t Mean Preferences
%A Jiang, Yiwen
%A Deng, Yang
%A Fong, Stephanie
%A Wang, Zimu
%A Shen, Yaling
%A Feng, Wei
%A Yang, Hongxi
%A Zhao, Xiangyu
%A Xu, Zhongxing
%A Mehta, Deval
%A Cheng, Xuelian
%A Ge, Zongyuan
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F jiang-etal-2026-vibe
%X Personalized Large Language Models (PLLMs) aim to tailor responses to individual users, where a central challenge is preference reasoning: inferring query-relevant preferences from user-related history. Existing benchmarks, however, largely assume that such preference can be retrieved from semantically related history. We study an underexplored but practically important regime, profile-preference conceptual misalignment (PRCM), where observable profile cues and query-specific preferences lie in different concept spaces, making semantic retrieval inconsistent for personalization. We introduce Vibe-Bench, a benchmark with two psychology-grounded tasks, 3,504 personas and 12,239 dialogues, including a manually verified gold test set, and requires cross-concept preference reasoning beyond surface semantic overlap. Experiments with several personalization methods show that current PLLMs largely rely on shallow semantic correlations and fail to acquire robust cross-concept mappings. These findings establish PRCM as a distinct failure regime in PLLMs and position Vibe-Bench as a focused testbed for advancing preference reasoning beyond semantic matching.
%U https://aclanthology.org/2026.findings-emnlp.1676/
%P 35771-35797
Markdown (Informal)
[VIBE-Bench: Evaluating Personalized Large Language Models When Profiles Don’t Mean Preferences](https://aclanthology.org/2026.findings-emnlp.1676/) (Jiang et al., Findings 2026)
ACL
- Yiwen Jiang, Yang Deng, Stephanie Fong, Zimu Wang, Yaling Shen, Wei Feng, Hongxi Yang, Xiangyu Zhao, Zhongxing Xu, Deval Mehta, Xuelian Cheng, and Zongyuan Ge. 2026. VIBE-Bench: Evaluating Personalized Large Language Models When Profiles Don’t Mean Preferences. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 35771–35797, Budapest, Hungary. Association for Computational Linguistics.