Home / AI Models / NP-GIANT
Model released image-feature-extraction apache-2.0

NP-GIANT

Details

Architecture ViT-Giant with Registers (~1B params)
Parameters ~1B
Base Model facebook/dinov2-with-registers-giant
Relation finetune
License apache-2.0

Neuropathology Vision Transformer using DINOv2-with-registers architecture. Key innovation is the hierarchically connected tiles strategy, which improves representation learning from whole slide images.

Results

MetricValue
Linear Probe Accuracy84.51%
KNN Accuracy87.20%

Outperforms UNI2 (77.5%), prov-gigapath (77.0%), UNI (76.7%) on the same dataset.