@inproceedings{cocchieri-etal-2026-lost,
title = "Lost in Choice: Evaluating {LLM}s Under Extreme Multiple-Choice Scaling",
author = "Cocchieri, Alessio and
Ragazzi, Luca and
Tagliavini, Giuseppe and
Moro, Gianluca",
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.328/",
pages = "6931--6963",
ISBN = "979-8-89176-462-0",
abstract = "Existing multiple-choice evaluations use small option sets and thus conflate several sources of failure when candidate spaces grow. We provide a controlled factorial decomposition of large-choice degradation across four axes: whether a question requires comparing its options, candidate cardinality (up to 1,000), spatial layout, and gold-answer position. Across six reasoning LLMs, comparative questions degrade far more than questions solvable without inspecting the options; dispersion amplifies this gap; and accuracy often collapses at late-middle and terminal positions rather than recovering in the classic lost-in-the-middle U-shape. We call this asymmetric, structure-dependent failure mode Lost in Choice. A replication on 500 real ToolBench tool-selection trajectories per condition reproduces the same hardest position on two additional model-task evaluations. Finally, an audit of reasoning traces yields a taxonomy of the heuristics and shortcuts models adopt as candidate spaces expand."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cocchieri-etal-2026-lost">
<titleInfo>
<title>Lost in Choice: Evaluating LLMs Under Extreme Multiple-Choice Scaling</title>
</titleInfo>
<name type="personal">
<namePart type="given">Alessio</namePart>
<namePart type="family">Cocchieri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Luca</namePart>
<namePart type="family">Ragazzi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Giuseppe</namePart>
<namePart type="family">Tagliavini</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gianluca</namePart>
<namePart type="family">Moro</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>Existing multiple-choice evaluations use small option sets and thus conflate several sources of failure when candidate spaces grow. We provide a controlled factorial decomposition of large-choice degradation across four axes: whether a question requires comparing its options, candidate cardinality (up to 1,000), spatial layout, and gold-answer position. Across six reasoning LLMs, comparative questions degrade far more than questions solvable without inspecting the options; dispersion amplifies this gap; and accuracy often collapses at late-middle and terminal positions rather than recovering in the classic lost-in-the-middle U-shape. We call this asymmetric, structure-dependent failure mode Lost in Choice. A replication on 500 real ToolBench tool-selection trajectories per condition reproduces the same hardest position on two additional model-task evaluations. Finally, an audit of reasoning traces yields a taxonomy of the heuristics and shortcuts models adopt as candidate spaces expand.</abstract>
<identifier type="citekey">cocchieri-etal-2026-lost</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.328/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>6931</start>
<end>6963</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Lost in Choice: Evaluating LLMs Under Extreme Multiple-Choice Scaling
%A Cocchieri, Alessio
%A Ragazzi, Luca
%A Tagliavini, Giuseppe
%A Moro, Gianluca
%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 cocchieri-etal-2026-lost
%X Existing multiple-choice evaluations use small option sets and thus conflate several sources of failure when candidate spaces grow. We provide a controlled factorial decomposition of large-choice degradation across four axes: whether a question requires comparing its options, candidate cardinality (up to 1,000), spatial layout, and gold-answer position. Across six reasoning LLMs, comparative questions degrade far more than questions solvable without inspecting the options; dispersion amplifies this gap; and accuracy often collapses at late-middle and terminal positions rather than recovering in the classic lost-in-the-middle U-shape. We call this asymmetric, structure-dependent failure mode Lost in Choice. A replication on 500 real ToolBench tool-selection trajectories per condition reproduces the same hardest position on two additional model-task evaluations. Finally, an audit of reasoning traces yields a taxonomy of the heuristics and shortcuts models adopt as candidate spaces expand.
%U https://aclanthology.org/2026.findings-emnlp.328/
%P 6931-6963
Markdown (Informal)
[Lost in Choice: Evaluating LLMs Under Extreme Multiple-Choice Scaling](https://aclanthology.org/2026.findings-emnlp.328/) (Cocchieri et al., Findings 2026)
ACL