Quantifying and Enabling the Interpretability of CLIP-like Models

VLM
Re-assign community