# Homoglyph-based Adversarial Perturbation of Introductory Computer Science Theory Problems

arXiv: https://arxiv.org/abs/2605.16286 · HTML: https://mlsanity.com/paper/2605.16286/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 10 of its 17 references that are in the map.
- First posted on arXiv 13 Apr 2026.

## Builds on (most cited first, 10 of 10)

- [MATH-Perturb: Benchmarking LLMs' Math Reasoning Abilities against Hard Perturbations](https://mlsanity.com/paper/2502.06453/) (2025)
- [Attacking Neural Text Detectors](https://mlsanity.com/paper/2002.11768/) (2020)
- [Neuro-Symbolic AI in 2024: A Systematic Review](https://mlsanity.com/paper/2501.05435/) (2025)
- [Large Language Models Are Neurosymbolic Reasoners](https://mlsanity.com/paper/2401.09334/) (2024)
- [SilverSpeak: Evading AI-Generated Text Detectors using Homoglyphs](https://mlsanity.com/paper/2406.11239/) (2024)
- [An Investigation of Robustness of LLMs in Mathematical Reasoning: Benchmarking with Mathematically-Equivalent Transformation of Advanced Mathematical Problems](https://mlsanity.com/paper/2508.08833/) (2025)
- [Impeding LLM-assisted Cheating in Introductory Programming Assignments via Adversarial Perturbation](https://mlsanity.com/paper/2410.09318/) (2024)
- [Visual Spoofing in content based spam detection](https://mlsanity.com/paper/2004.05265/) (2020)
- [The Lazy Student's Dream: ChatGPT Passing an Engineering Course on Its Own](https://mlsanity.com/paper/2503.05760/) (2025)
- [LLM Use, Cheating, and Academic Integrity in Software Engineering Education](https://mlsanity.com/paper/2603.17060/) (2026)

## Abstract (arXiv)

Different AI tools such as ChatGPT, Gemini, and Claude are becoming very popular. Although they are helpful for many day-to-day tasks, they can be used in unexpected ways. For example, the learning objectives of a course may not be achieved if students use these tools to solve their homework problems. This paper proposes a simple method to address this issue in the lazy student model. The method uses homoglyph-based adversarial perturbation to first modify the question without changing the semantic meaning of the question. Then a few characters are perturbed by their homoglyphs. Our experimental result shows the theoretical problems of introductory computer science courses can be effectively perturbed. We also propose an interactive tool to conveniently use our method.