Unleashing the Power of Meta-tuning for Few-shot Generalization Through Sparse Interpolated Experts

Benchmark Model Rank Results
few-shot-image-classification-on-meta-datasetSMAT (DINO-VIT-Base-16-224)#1Accuracy: 85.27