# Simulating Novice Students Using Machine Unlearning and Relearning in Large Language Models

arXiv: https://arxiv.org/abs/2603.26142 · HTML: https://mlsanity.com/paper/2603.26142/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Machine unlearning, one of 12 directions in LLM safety and security.
- Builds on 13 of its 29 references that are in the map.
- First posted on arXiv 27 Mar 2026.

## Placed in

- LLM safety and security / [Machine unlearning](https://mlsanity.com/safety/machine-unlearning/)

## Builds on (most cited first, 13 of 13)

- [Machine Unlearning](https://mlsanity.com/paper/1912.03817/) (2019)
- [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023)
- [DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models](https://mlsanity.com/paper/2512.02556/) (2025)
- [Rethinking Machine Unlearning for Large Language Models](https://mlsanity.com/paper/2402.08787/) (2024)
- [Position: LLM Unlearning Benchmarks are Weak Measures of Progress](https://mlsanity.com/paper/2410.02879/) (2024)
- [Generative Students: Using LLM-Simulated Student Profiles to Support Question Item Evaluation](https://mlsanity.com/paper/2405.11591/) (2024)
- [Teach AI How to Code: Using Large Language Models as Teachable Agents for Programming Education](https://mlsanity.com/paper/2309.14534/) (2023)
- [Revisiting Who's Harry Potter: Towards Targeted Unlearning from a Causal Intervention Perspective](https://mlsanity.com/paper/2407.16997/) (2024)
- [How to Teach Programming in the AI Era? Using LLMs as a Teachable Agent for Debugging](https://mlsanity.com/paper/2310.05292/) (2023)
- [ReLearn: Unlearning via Learning for Large Language Models](https://mlsanity.com/paper/2502.11190/) (2025)
- [Simulated Students in Tutoring Dialogues: Substance or Illusion?](https://mlsanity.com/paper/2601.04025/) (2026)
- [Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills](https://mlsanity.com/paper/2506.12963/) (2025)
- [Learning-by-teaching with ChatGPT: The effect of teachable ChatGPT agent on programming education](https://mlsanity.com/paper/2412.15226/) (2024)

## Built on by (most cited first, 3 of 3)

- [CustomerSim: Benchmarking and Aligning Multimodal Language Models as Retail User Simulators](https://mlsanity.com/paper/2605.08334/) (2026)
- [Simulating Disengaged Students to Evaluate LLM-based Tutors](https://mlsanity.com/paper/2609.12331/) (2026)
- [Sherpa: Teaching LLMs to Teach Adaptively](https://mlsanity.com/paper/2610.08778/) (2026)

## Abstract (arXiv)

Student simulation can support learning-by-teaching pedagogy where human students (as tutors) teach AI-simulated novice students (as tutees). Recent research often relies on prompt engineering with large language models (LLMs) to simulate novice student behaviour, but it is difficult to keep the AI-simulated student at a stable novice knowledge level. A key reason is that many LLMs are trained to be broadly capable, so even when prompted to "act like a novice," the LLMs can still produce expert-level explanations during the learning-by-teaching interaction process. As a result, the AI-simulated student may drift beyond the intended knowledge level, reducing the credibility of the simulation for studying learning-by-teaching processes. Thus, we propose a knowledge-level simulation approach based on machine unlearning. We investigate this approach using a dataset of multiple-choice questions on Python programming concepts. We apply machine unlearning to transform a knowledgeable LLM into a novice-level AI student (i.e., teachable agent), then evaluate whether the teachable agent can relearn targeted knowledge components through learning-by-teaching dialogue interactions. Finally, we analyse the dialogue logs to characterise how the agent's behaviour changes over time, including its question asking, error patterns, and responsiveness to instruction. The results show that (1) unlearning produces simulated student agents with more novice-like responses than prompt-only baselines, (2) the agents recover a measurable portion of the unlearned knowledge under structured exposure, and (3) dialogue analyses reveal identifiable trajectories of conceptual change and teaching moves that predict learning recovery.