Is there a transparent way to interleave text and image tokens to test in-context learning abilities for the model?
· Sign up or log in to comment
