Skip to main content

Mithridatium

Overview

Mithridatium is a set of tools that translates research on detecting and prevention of AI poisoning attacks into practical software solutions.

Basic Information

Core Dev Team

  • Client Dr. Reza Tourani
  • Track: Client-driven Product
  • Current Tech Lead: Pelumi Oluwategbe github

  • Developers:

    • Payton Guffey github
    • Gustavo Lucca github
    • Will Phoenix github

Technical Information

Additional Information

  • Start Date: August 2025
  • Technologies Used:
    • Python
    • AI/ML (poisoning attack detection and prevention)
  • License: MIT
  • Code of Conduct: CODE_OF_CONDUCT.md

Development Priorities

  • Research and implement AI poisoning attack detection techniques
  • Translate academic research into practical, reusable software tools
  • Build testing frameworks for evaluating model robustness against poisoning
  • Document detection and prevention best practices

Get Involved

If you would like to contribute to this project, please visit our GitHub page to create your own issues or pull requests.