How to draw the figures in hacl？ #5
Description
We have recently been exploring methods to reduce the modality gap in multimodal large models, but have been struggling to determine the most effective way to validate the efficacy of our approaches. Fortunately, we came across your excellent paper, HACL, and were particularly inspired by the image mentioned above. We are reaching out to inquire about how to create such a visual representation.
Looking forward to your reply!