Fetching the paper…

Rethinking Misalignment in Vision-Language Model Adaptation from a Causal Perspective · Around