prithivMLmods/AIorNot-SigLIP2
Full fine-tune (fewer layers) vs google/siglip2-base-patch16-224. prithivMLmods/AIorNot-SigLIP2 matches part of google/siglip2-base-patch16-224 (208 of its 408 tensors; the rest of google/siglip2-base-patch16-224 has no counterpart). On the shared part: Fine-tune: 92% of the weights changed, across 12 of 12 layers.
Tensors changed: Embeddings 3/3, Norms 50/52, MLP 48/52, Attention 96/100, other 0/1