@inproceedings{dlugosz-etal-2026-importance,
title = "The Importance of Being Statistically Earnest: A Critical Re-evaluation of {GSM}-Symbolic",
author = "D{\l}ugosz, Dominika Agnieszka and
Oliveira, Arlindo L. and
D{\'i}az-Rodr{\'i}guez, Natalia",
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.1487/",
pages = "33984--34024",
ISBN = "979-8-89176-459-0",
abstract = "The GSM-Symbolic benchmark (Mirzadeh et al., 2025) reported consistent performance drops across 25 Large Language Models (LLMs) when tested on template-generated variants of GSM8K problems, concluding that the models lack genuine reasoning capabilities. We argue that this conclusion rests on shaky statistical ground. Re-evaluating 20 open-weight models using bootstrapped Generalised Linear Mixed Models with per-question random effects, we find that only 8 exhibit statistically significant performance changes under the original prompt format. Moreover, we identify a previously unacknowledged factor: the distribution of integers in problem texts of the main GSM-Symbolic dataset is systematically shifted towards larger values relative to the original GSM8K (K-S statistic = 0.12, p {\ensuremath{<}} 0.001), contradicting the original authors' claims. Controlling for this large-number effect accounts for significance in half of the remaining cases. Among models with statistically significant performance deltas, we identify distinct, model-specific behavioural failure profiles {---} including fragility of variable binding, arithmetic limitations, and dual-task interference {---} underscoring that blanket claims about LLM reasoning risk being both statistically premature and mechanistically misleading."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="dlugosz-etal-2026-importance">
<titleInfo>
<title>The Importance of Being Statistically Earnest: A Critical Re-evaluation of GSM-Symbolic</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dominika</namePart>
<namePart type="given">Agnieszka</namePart>
<namePart type="family">Długosz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Arlindo</namePart>
<namePart type="given">L</namePart>
<namePart type="family">Oliveira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Natalia</namePart>
<namePart type="family">Díaz-Rodríguez</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>The GSM-Symbolic benchmark (Mirzadeh et al., 2025) reported consistent performance drops across 25 Large Language Models (LLMs) when tested on template-generated variants of GSM8K problems, concluding that the models lack genuine reasoning capabilities. We argue that this conclusion rests on shaky statistical ground. Re-evaluating 20 open-weight models using bootstrapped Generalised Linear Mixed Models with per-question random effects, we find that only 8 exhibit statistically significant performance changes under the original prompt format. Moreover, we identify a previously unacknowledged factor: the distribution of integers in problem texts of the main GSM-Symbolic dataset is systematically shifted towards larger values relative to the original GSM8K (K-S statistic = 0.12, p \ensuremath< 0.001), contradicting the original authors’ claims. Controlling for this large-number effect accounts for significance in half of the remaining cases. Among models with statistically significant performance deltas, we identify distinct, model-specific behavioural failure profiles — including fragility of variable binding, arithmetic limitations, and dual-task interference — underscoring that blanket claims about LLM reasoning risk being both statistically premature and mechanistically misleading.</abstract>
<identifier type="citekey">dlugosz-etal-2026-importance</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1487/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>33984</start>
<end>34024</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The Importance of Being Statistically Earnest: A Critical Re-evaluation of GSM-Symbolic
%A Długosz, Dominika Agnieszka
%A Oliveira, Arlindo L.
%A Díaz-Rodríguez, Natalia
%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 dlugosz-etal-2026-importance
%X The GSM-Symbolic benchmark (Mirzadeh et al., 2025) reported consistent performance drops across 25 Large Language Models (LLMs) when tested on template-generated variants of GSM8K problems, concluding that the models lack genuine reasoning capabilities. We argue that this conclusion rests on shaky statistical ground. Re-evaluating 20 open-weight models using bootstrapped Generalised Linear Mixed Models with per-question random effects, we find that only 8 exhibit statistically significant performance changes under the original prompt format. Moreover, we identify a previously unacknowledged factor: the distribution of integers in problem texts of the main GSM-Symbolic dataset is systematically shifted towards larger values relative to the original GSM8K (K-S statistic = 0.12, p \ensuremath< 0.001), contradicting the original authors’ claims. Controlling for this large-number effect accounts for significance in half of the remaining cases. Among models with statistically significant performance deltas, we identify distinct, model-specific behavioural failure profiles — including fragility of variable binding, arithmetic limitations, and dual-task interference — underscoring that blanket claims about LLM reasoning risk being both statistically premature and mechanistically misleading.
%U https://aclanthology.org/2026.emnlp-main.1487/
%P 33984-34024
Markdown (Informal)
[The Importance of Being Statistically Earnest: A Critical Re-evaluation of GSM-Symbolic](https://aclanthology.org/2026.emnlp-main.1487/) (Długosz et al., EMNLP 2026)
ACL