List your company
ResearchFormerly WizWand

Node Classification on German

74.9AccuracyFairAdj
45.88453.41760.9568.483Oct 27, 2025Nov 17, 2025Dec 8, 2025Dec 29, 2025Jan 19, 2026Feb 9, 2026Mar 2, 2026
Updated 4mo ago

Evaluation results

MethodLinks
FairAdjFramework=Graph Contrastive Learning2026.0374.9—0.990.78——
FairGDiffFramework=Graph Contrastive Learning2026.0374.27—0.120.17——
FairGenFramework=Graph Contrastive Learning2026.0374.04—0.020.05——
GraphairFramework=Graph Contrastive Learning2026.0373.69—4.894.58——
DiffusionFramework=Graph Contrastive Learning2026.0373.02—5.946.45——
FairDropFramework=Graph Contrastive Learning, p=0.52026.0372.84—9.737.72——
GCLFramework=Graph Contrastive Learning2026.0372.58—18.8215.53——
FairDropFramework=Graph Contrastive Learning, p=0.752026.0371.78—11.1311.94——
ADPromptPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1065.5——2.14—2.62
ADPromptPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1065——2.34—2.37
ClassifierPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1064.5——4.52—2.63
FPromptPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1064.5——3.35—3.43
ADPromptPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1064.33——1.18—1.32
Cls. onlyPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1063.33——7.23—9.6
ADPromptPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1062.17——2.22—2.32
Self-proPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1062——2.96—5.51
Self-proPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1061.83——8.95—5.8
Adv. learningPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1061.8——0.82—6.39
FPromptPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1061.67——6.13—4.57
GraphPromptPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1061.5——2.21—2.51
GPF+Pre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1061——8.49—4.49
GPFPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1060.25——6.07—3.53
GraphPromptPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1060.17——6.44—5.63
Adv. learningPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1060.17——5.2—3.71
GPF+Pre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1059.33——5.58—4.02
GraphPromptPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1059.17——3.37—3.06
FPromptPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1058.67——3.47—3.73
Cls. onlyPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1058.5——8.5—7.29
GraphPromptPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1058.49——5.54—4.96
GPFPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1058.17——1.7—6.34
Adv. learningPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1057.83——2.58—4.06
Adv. learningPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1057——8.47—9.78
GPF+Pre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1056.83——4.78—5.07
Cls. onlyPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1055.5——5.72—7.09
FPromptPre-train Strategy=InfoMax, Shot Count=50-shot, Backbone=2-layer GCN2025.1052.5——4.22—5.11
GPF+Pre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1051——6.5—6.97
Self-proPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1050.67——2.71—3.57
GPFPre-train Strategy=GAE, Shot Count=50-shot, Backbone=2-layer GCN2025.1050.33——7.38—4.68
GPFPre-train Strategy=GraphCL, Shot Count=50-shot, Backbone=2-layer GCN2025.1050.07——7.09—3.73
Self-proPre-train Strategy=BGRL, Shot Count=50-shot, Backbone=2-layer GCN2025.1047——5.74—9.2
DiffusionFramework=Graph Contrastive Learning2026.03————3.4413—
FairAdjFramework=Graph Contrastive Learning2026.03————8.2017—
FairDropFramework=Graph Contrastive Learning, dropout rate=0.52026.03————2.084—
FairDropFramework=Graph Contrastive Learning, dropout rate=0.752026.03————0.79—
FairGDiffFramework=Graph Contrastive Learning2026.03————8.7715—
FairGenFramework=Graph Contrastive Learning2026.03————8.8193—
GCLFramework=Graph Contrastive Learning2026.03————0.0002—
GraphairFramework=Graph Contrastive Learning2026.03————4.5838—