Robustness
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
42.5mCE
30
May 21, 2026
64.11Clean Accuracy
24
Jul 7, 2026
61.01R-Bench Dis Metric
13
Jun 1, 2026
87.97RR
9
Feb 26, 2026
40.78mCE
9
Feb 26, 2026
58.5Error Rate
8
Feb 26, 2026
44.3Error Rate
8
Feb 26, 2026
77.6Mean Corruption Error (mCE)
5
May 21, 2026
66.69Robustness Accuracy
4
Jun 19, 2026
0.3Hamming Distance (x100)
4
Feb 26, 2026
1.2Hamming Distance (x100)
4
Feb 26, 2026
1.1Hamming Distance (x100)
4
Feb 26, 2026
63.11GaaA Win Rate
3
Apr 10, 2026
47.9Average Score
3
Feb 26, 2026
39.06PGD Robustness
2
Mar 13, 2026