List your company
ResearchFormerly WizWand

Model Merging on Sustainability to large-scale 8 tasks

94.8Avg Normalized AccuracyDOGE AM
72.85678.55384.2589.947Aug 5, 2025
Updated 6mo ago

Evaluation results

MethodLinks
DOGE AMBackbone=ViT-B/32, Method Category=Test-Time Adaption2025.0894.8
RegMean++Backbone=ViT-B/32, Method Category=Training-Free2025.0893.1
TSV-MBackbone=ViT-B/32, Method Category=Data-Free2025.0891.6
Iso-CTSBackbone=ViT-B/32, Method Category=Data-Free2025.0891.6
Iso-CBackbone=ViT-B/32, Method Category=Data-Free2025.0891.3
Layer-wise AdaMergingBackbone=ViT-B/32, Method Category=Test-Time Adaption2025.0891.2
RegMeanBackbone=ViT-B/32, Method Category=Training-Free2025.0890.9
DOGE TABackbone=ViT-B/32, Method Category=Data-Free2025.0889
TIES-MergingBackbone=ViT-B/32, Method Category=Data-Free2025.0879.6
Fisher MergingBackbone=ViT-B/32, Method Category=Training-Free2025.0878.2
Task ArithmeticBackbone=ViT-B/32, Method Category=Data-Free2025.0874.3
Model SoupsBackbone=ViT-B/32, Method Category=Data-Free2025.0873.7