Women In AI Safety Hackathon

Challenge Tracks

🎛 Mechanistic Interpretability (sponsored by Goodfire)

- Understand AI model internals with state-of-the-art techniques

- Develop model steering and editing methods

- Build interactive tools to visualize and interpret AI behavior

🌍Public Education (sponsored by BlueDot Impact)

- Create AI safety games for non-technical learners

- Design a mentorship platform for AI safety newcomers

- Build tools for self-assessment and learning paths

🧠 Social Sciences & AI

- Explore AI cognition through psychology and behavioral science

- Analyze AI’s impact on human society, ethics, and governance

- Contribute to interdisciplinary research with technical and non-technical project

Signup: https://docs.google.com/forms/d/e/1FAIpQLSfkzPYmNGjOJKRmMsAFVxJksqRMf-f3JBzTOoAWum9PUtVVZA/viewform?usp=header