Chameleon
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Chameleon
40.044PSNR
8
Chameleon multiscale (ten pre-defined splits)
0.131NMSE
8
Chameleon
0.99AUC
8
Chameleon Directed (test)
78.14Accuracy
8
Chameleon overall
53.21Accuracy
8
Chameleon high-degree nodes, upper 33%
54.53Accuracy
8
Chameleon low-degree nodes, lower 33%
54.48Accuracy
8
Chameleon
0Sharpness
7
Chameleon H = .26
56.1Accuracy
7
Chameleon h=0.23 (100 labeled nodes)
37.2Micro-F1
7
Chameleon h=0.23 (75 labeled nodes)
42.8Micro-F1
7
Chameleon h=0.23 (50 labeled nodes)
35.5Micro-F1
7
Chameleon (average over 5 different splits)
52.87Accuracy
7
CHAMELEON
0.018MAE
6
Chameleon (test)
0.136Normalized Test MSE
6
CHAMELEON CF-COD paired benchmark (test)
93.4Do
5
Chameleon
84.18AUROC
5
Chameleon 10% (train/val)
35.4Accuracy
4
Chameleon Hard to Unlearn
69.4ToU
4
Chameleon Random
42.6ToU
4
Chameleon Easy to Unlearn
84.4ToU Score
4
Chameleon unobserved (test)
32.89Mean Accuracy
4
Chameleon observed (test)
42.23Mean Accuracy
4
Chameleon (test)
92Test AUC
4
Chameleon (ev)
71.5Identity Score
3