> Markdown version of [/@aldan-creo](https://www.wearedevelopers.com/@aldan-creo). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Aldan Creo
NLP researcher at UC San Diego exploring AI vulnerabilities and knowledge graphs.

## About

Aldan Creo combines natural language processing and knowledge graphs in his research. Currently an MS student at UC San Diego, his computer science background spans programs in Spain, France, and Switzerland.

He spends his time figuring out how AI systems break and how to secure them. His recent work explores how simple homoglyphs can manipulate tokenization, allowing text to completely evade AI detectors.

When he isn't researching vulnerabilities, Aldan contributes to open-source projects and builds student tech communities.


## Upcoming Sessions

### World Congress 2026 North America · September 23, 2026
San José, CA
- [Attacking LLM Detectors with Homoglyph-Based Attacks](https://www.wearedevelopers.com/events/world-congress-2026-north-america/sessions/1949-attacking-llm) · 120 min





## Links

- [LinkedIn](https://linkedin.com/in/acmcmc)