@inproceedings{yuan-etal-2026-codae,
title = "{C}o{DAE}: Adapting Large Language Models for Education via Chain-of-Thought Data Augmentation",
author = {Yuan, Shuzhou and
LaCroix, Willliam and
Ghoshal, Hardik and
Nie, Ercong and
F{\"a}rber, Michael},
editor = "Piperidis, Stelios and
Bel, N{\'u}ria and
van den Heuvel, Henk and
Ide, Nancy and
Krek, Simon and
Toral, Antonio",
booktitle = "Proceedings of the Fifteenth Language Resources and Evaluation Conference",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "ELRA Language Resource Association",
url = "https://aclanthology.org/2026.lrec-1.837/",
doi = "10.63317/447hbzyohzyf",
pages = "10677--10687",
abstract = "Large Language Models (LLMs) are increasingly employed as AI tutors in education due to their scalability and potential for personalized instruction. However, off-the-shelf LLMs often underperform in educational settings, exhibiting limitations such as providing answers too readily, failing to adapt their responses to students' uncertainty, and remaining susceptible to emotionally manipulative prompts. To address these challenges, we introduce CoDAE, a framework that adapts LLMs for educational use through Chain-of-Thought (CoT) data augmentation. We collect real-world dialogues between students and a ChatGPT-based tutor and enrich them using CoT prompting to promote step-by-step reasoning and pedagogically aligned guidance. Furthermore, we design targeted dialogue cases to explicitly mitigate three key limitations: over-compliance, low response adaptivity, and threat vulnerability. We fine-tune four open-source LLMs on different variants of the augmented datasets and evaluate them in simulated educational scenarios using both automatic metrics and LLM-as-a-judge assessments. Our results show that models fine-tuned with CoDAE deliver more pedagogically appropriate guidance, promote student reflection and more effectively prevent premature answer disclosure."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yuan-etal-2026-codae">
<titleInfo>
<title>CoDAE: Adapting Large Language Models for Education via Chain-of-Thought Data Augmentation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Shuzhou</namePart>
<namePart type="family">Yuan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Willliam</namePart>
<namePart type="family">LaCroix</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hardik</namePart>
<namePart type="family">Ghoshal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ercong</namePart>
<namePart type="family">Nie</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Michael</namePart>
<namePart type="family">Färber</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifteenth Language Resources and Evaluation Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Stelios</namePart>
<namePart type="family">Piperidis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Núria</namePart>
<namePart type="family">Bel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henk</namePart>
<namePart type="family">van den Heuvel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nancy</namePart>
<namePart type="family">Ide</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Simon</namePart>
<namePart type="family">Krek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="family">Toral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resource Association</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Large Language Models (LLMs) are increasingly employed as AI tutors in education due to their scalability and potential for personalized instruction. However, off-the-shelf LLMs often underperform in educational settings, exhibiting limitations such as providing answers too readily, failing to adapt their responses to students’ uncertainty, and remaining susceptible to emotionally manipulative prompts. To address these challenges, we introduce CoDAE, a framework that adapts LLMs for educational use through Chain-of-Thought (CoT) data augmentation. We collect real-world dialogues between students and a ChatGPT-based tutor and enrich them using CoT prompting to promote step-by-step reasoning and pedagogically aligned guidance. Furthermore, we design targeted dialogue cases to explicitly mitigate three key limitations: over-compliance, low response adaptivity, and threat vulnerability. We fine-tune four open-source LLMs on different variants of the augmented datasets and evaluate them in simulated educational scenarios using both automatic metrics and LLM-as-a-judge assessments. Our results show that models fine-tuned with CoDAE deliver more pedagogically appropriate guidance, promote student reflection and more effectively prevent premature answer disclosure.</abstract>
<identifier type="citekey">yuan-etal-2026-codae</identifier>
<identifier type="doi">10.63317/447hbzyohzyf</identifier>
<location>
<url>https://aclanthology.org/2026.lrec-1.837/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>10677</start>
<end>10687</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CoDAE: Adapting Large Language Models for Education via Chain-of-Thought Data Augmentation
%A Yuan, Shuzhou
%A LaCroix, Willliam
%A Ghoshal, Hardik
%A Nie, Ercong
%A Färber, Michael
%Y Piperidis, Stelios
%Y Bel, Núria
%Y van den Heuvel, Henk
%Y Ide, Nancy
%Y Krek, Simon
%Y Toral, Antonio
%S Proceedings of the Fifteenth Language Resources and Evaluation Conference
%D 2026
%8 May
%I ELRA Language Resource Association
%C Palma de Mallorca, Spain
%F yuan-etal-2026-codae
%X Large Language Models (LLMs) are increasingly employed as AI tutors in education due to their scalability and potential for personalized instruction. However, off-the-shelf LLMs often underperform in educational settings, exhibiting limitations such as providing answers too readily, failing to adapt their responses to students’ uncertainty, and remaining susceptible to emotionally manipulative prompts. To address these challenges, we introduce CoDAE, a framework that adapts LLMs for educational use through Chain-of-Thought (CoT) data augmentation. We collect real-world dialogues between students and a ChatGPT-based tutor and enrich them using CoT prompting to promote step-by-step reasoning and pedagogically aligned guidance. Furthermore, we design targeted dialogue cases to explicitly mitigate three key limitations: over-compliance, low response adaptivity, and threat vulnerability. We fine-tune four open-source LLMs on different variants of the augmented datasets and evaluate them in simulated educational scenarios using both automatic metrics and LLM-as-a-judge assessments. Our results show that models fine-tuned with CoDAE deliver more pedagogically appropriate guidance, promote student reflection and more effectively prevent premature answer disclosure.
%R 10.63317/447hbzyohzyf
%U https://aclanthology.org/2026.lrec-1.837/
%U https://doi.org/10.63317/447hbzyohzyf
%P 10677-10687
Markdown (Informal)
[CoDAE: Adapting Large Language Models for Education via Chain-of-Thought Data Augmentation](https://aclanthology.org/2026.lrec-1.837/) (Yuan et al., LREC 2026)
ACL