The checkpoint's projection_dim is 2048 and all three projection heads output 2048, as stated in the paper (Sec. 3, z in R^2048); commit 23107503 had changed the card to 512. Also note that image patches and expression profiles need their own input preparation, with a pointer to the GitHub README examples.
Correct README: 512-d shared space, fix package import + GitHub URL, fill in verified eval numbers