SAEV_Anole-7b_OBELICS is a sparse autoencoder with Vision (SAE-V) based on Chameleon model and trained on OBELICS dataset.
