Learning to Defer
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
70.16Accuracy
49
May 20, 2026
98.9Coverage
36
May 1, 2026
100Coverage
25
Apr 13, 2026
91.9System Accuracy
21
Apr 13, 2026
41.18Error
20
Feb 26, 2026
18.09Error Rate
20
Feb 26, 2026
41.23Error Rate
20
Feb 26, 2026
41.32Error Rate
20
Feb 26, 2026
15.1Error Rate
20
Feb 26, 2026
18.11Error Rate
20
Feb 26, 2026
96.1System Accuracy
18
Apr 13, 2026
71.8Deferral Ratio (%) - Expert 1
12
May 1, 2026
83.27Deferral Ratio (Expert 1)
12
May 1, 2026
71.31Deferral Ratio (Expert 1)
12
May 1, 2026
0.427Deferral Loss
12
May 1, 2026
0.106Deferral Loss
12
May 1, 2026
76SAC
12
Mar 4, 2026
72.36AUACC
10
Apr 2, 2026
76.93AUACC
10
Apr 2, 2026
79.7AU Accuracy
10
Apr 2, 2026
76SAC
9
Mar 4, 2026
88SAC
8
Mar 4, 2026
89SAC
8
Mar 4, 2026
86SAC
8
Mar 4, 2026
0.058Deferral Loss
6
May 1, 2026