AI

Confirming Claims of Superposition and Adversarial Examples in Toy Models

A replication of core claims about superposition in toy models, with an interactive animation of a PGD adversarial attack unfolding in the model’s latent space.