PhD candidate at the CIRCLE and Safe AI Lab research labs at Imperial, researching mechanistic interpretability and neural network robustness.
Research engineer at Safe Intelligence.
Contact me at edward.stevinson@hotmail.com