AI safety research nonprofit founded by Adam Gleave and Karl Berzins, based in Berkeley. Works on adversarial robustness, model evaluation, interpretability, and alignment, and convenes the AI safety research community.
See something inaccurate or outdated?
