Scale Alone Does not Improve Mechanistic Interpretability in Vision Models

Devs

Scale Alone Does not Improve Mechanistic Interpretability in Vision Models | Read Paper on Bytez