Sahil Raut

Mumbai, India

Bio

I obsess over red-teaming frontier models (this is something that always keeps me awake at night!) and currently researching independently on various topics such as Alignment, Evals and data, Interpretability and SAE features.

Links

Personal Website: https://sahilraut.com/
LinkedIn: LinkedIn
Twitter / X: -
LessWrong: https://www.lesswrong.com/users/sahilraut
EA Forum: -

Projects

Benchmark for Reversal and Unlearning of Harmful Fine-Tuning

Team Member

Grants

No grants recorded.