@inproceedings{al-nazi-etal-2026-dagger,
title = "{\textdagger}{DAGGER}: Distractor-Aware Graph Generation for Executable Reasoning in Math Problems",
author = "Al Nazi, Zabir and
Roy Dipta, Shubhashis and
Kar, Sudipta",
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.2261/",
pages = "48423--48446",
ISBN = "979-8-89176-462-0",
abstract = "Chain-of-Thought (CoT) prompting is widely adopted for mathematical problem solving, including in low-resource languages, yet its robustness under irrelevant context - a known vulnerability in English - remains unexamined in such settings. To systematically probe this challenge, we introduce DISTRACTMATH-BN, a Bangla benchmark that augments MGSM and MSVAMP with semantically coherent but computationally irrelevant information. Evaluating seven models ranging from 3B to 12B parameters, we observe substantial performance degradation under distractors: standard models drop up to 41 points, while reasoning-specialized models decline 14-20 points despite consuming 5{\texttimes} more tokens. We propose {\textdagger}DAGGER, which reformulates mathematical problem solving as executable computational graph generation with dedicated modeling of distractor nodes. Fine-tuning Gemma-3 models using Supervised Fine-Tuning followed by Group Relative Policy Optimization achieves comparable weighted accuracy on augmented benchmarks while using 89{\%} fewer tokens than reasoning models. Importantly, this robustness emerges without explicit training on distractor-augmented examples. Our results suggest that enforcing structured intermediate representations can improve robustness and inference efficiency in mathematical understanding compared to free-form reasoning, particularly in noisy, low-resource settings."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="al-nazi-etal-2026-dagger">
<titleInfo>
<title>†DAGGER: Distractor-Aware Graph Generation for Executable Reasoning in Math Problems</title>
</titleInfo>
<name type="personal">
<namePart type="given">Zabir</namePart>
<namePart type="family">Al Nazi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shubhashis</namePart>
<namePart type="family">Roy Dipta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sudipta</namePart>
<namePart type="family">Kar</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>Chain-of-Thought (CoT) prompting is widely adopted for mathematical problem solving, including in low-resource languages, yet its robustness under irrelevant context - a known vulnerability in English - remains unexamined in such settings. To systematically probe this challenge, we introduce DISTRACTMATH-BN, a Bangla benchmark that augments MGSM and MSVAMP with semantically coherent but computationally irrelevant information. Evaluating seven models ranging from 3B to 12B parameters, we observe substantial performance degradation under distractors: standard models drop up to 41 points, while reasoning-specialized models decline 14-20 points despite consuming 5× more tokens. We propose †DAGGER, which reformulates mathematical problem solving as executable computational graph generation with dedicated modeling of distractor nodes. Fine-tuning Gemma-3 models using Supervised Fine-Tuning followed by Group Relative Policy Optimization achieves comparable weighted accuracy on augmented benchmarks while using 89% fewer tokens than reasoning models. Importantly, this robustness emerges without explicit training on distractor-augmented examples. Our results suggest that enforcing structured intermediate representations can improve robustness and inference efficiency in mathematical understanding compared to free-form reasoning, particularly in noisy, low-resource settings.</abstract>
<identifier type="citekey">al-nazi-etal-2026-dagger</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.2261/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>48423</start>
<end>48446</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T †DAGGER: Distractor-Aware Graph Generation for Executable Reasoning in Math Problems
%A Al Nazi, Zabir
%A Roy Dipta, Shubhashis
%A Kar, Sudipta
%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 al-nazi-etal-2026-dagger
%X Chain-of-Thought (CoT) prompting is widely adopted for mathematical problem solving, including in low-resource languages, yet its robustness under irrelevant context - a known vulnerability in English - remains unexamined in such settings. To systematically probe this challenge, we introduce DISTRACTMATH-BN, a Bangla benchmark that augments MGSM and MSVAMP with semantically coherent but computationally irrelevant information. Evaluating seven models ranging from 3B to 12B parameters, we observe substantial performance degradation under distractors: standard models drop up to 41 points, while reasoning-specialized models decline 14-20 points despite consuming 5× more tokens. We propose †DAGGER, which reformulates mathematical problem solving as executable computational graph generation with dedicated modeling of distractor nodes. Fine-tuning Gemma-3 models using Supervised Fine-Tuning followed by Group Relative Policy Optimization achieves comparable weighted accuracy on augmented benchmarks while using 89% fewer tokens than reasoning models. Importantly, this robustness emerges without explicit training on distractor-augmented examples. Our results suggest that enforcing structured intermediate representations can improve robustness and inference efficiency in mathematical understanding compared to free-form reasoning, particularly in noisy, low-resource settings.
%U https://aclanthology.org/2026.findings-emnlp.2261/
%P 48423-48446
Markdown (Informal)
[†DAGGER: Distractor-Aware Graph Generation for Executable Reasoning in Math Problems](https://aclanthology.org/2026.findings-emnlp.2261/) (Al Nazi et al., Findings 2026)
ACL