PaliGemma2 LoRA finetuned on VQAv2
Identify key points in an image
Vision Transformer Attention Visualization