CoDAE: Adapting Large Language Models for Education via Chain-of-Thought Data Augmentation

TL;DR

This paper introduces CoDAE, a data augmentation framework using Chain-of-Thought prompting to adapt large language models for educational purposes, improving their reasoning, adaptivity, and resistance to manipulation.

Contribution

It presents a novel CoT-based data augmentation method to fine-tune LLMs for education, addressing key limitations like over-compliance and vulnerability.

Findings

Enhanced pedagogical guidance in fine-tuned models

Improved reasoning and response adaptivity

Increased resistance to manipulative prompts

Abstract

Large Language Models (LLMs) are increasingly employed as AI tutors due to their scalability and potential for personalized instruction. However, off-the-shelf LLMs often underperform in educational settings: they frequently reveal answers too readily, fail to adapt their responses to student uncertainty, and remain vulnerable 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, better support reasoning processes, and effectively resist premature answer disclosure.