Announcement_1

Our paper on adversarial robustness of concept representations in sparse autoencoders was accepted to EACL 2026!