Probing Relational Reasoning in Pretrained ViTs
We explore whether vision transformers can perform abstract reasoning tasks using by solving puzzles. CLIP vs. DiNOv3
We explore whether vision transformers can perform abstract reasoning tasks using by solving puzzles. CLIP vs. DiNOv3
Log in or sign up for Devpost to join the conversation.