Image Classification on CIFAR-10 non-IID s=2 benchmark leaderboard · SOTA2 Research

Image Classification on CIFAR-10 non-IID s=2

91.82Test AccuracyCD²-pFed

Evaluation Results

Method	Links
CD²-pFed2022.04		91.82
LG-Fed2022.04		91.77
Ensemble2025.06		89.1
Local Train2022.04		87.92
KF-Linear Conductance2025.06		86.6
KF-Linear DeepLIFT2025.06		86.5
HF-Linear Conductance2025.06		86.4
HF-Linear DeepLIFT2025.06		86.3
LP2025.06		85.5
KF-Linear Uniform2025.06		85.5
HF-Gradient DeepLIFT2025.06		84.9
KF-Gradient Conductance2025.06		84.9
HF-Gradient Conductance2025.06		84.8
KF-Gradient DeepLIFT2025.06		84.7
KF-Gradient Uniform2025.06		84.3
Individual Models2025.06		83.8
FedPer2022.04		83.29
HF-Linear Uniform2025.06		76.7
Git Re-Basin DeepLIFT2025.06		76.6
HF-Gradient Uniform2025.06		76.3
Git Re-Basin Conductance2025.06		73.8
CD²-pFed2022.04		61.31
LG-Fed2022.04		60.79
FedAvg2022.04		58.99
FedAvg2022.04		58.99
FedPer2022.04		57.77
Git Re-Basin Uniform2025.06		55.7
OTF Uniform2025.06		49.6
OTF Conductance2025.06		40.6
OTF DeepLIFT2025.06		39.4
KD2025.06		38.4
Vanilla Averaging2025.06		11.5
Local Train2022.04		10.03