Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Toxicity Classification on CivilComments
Loading...
92.6
Average Accuracy
ERM
86.672
88.211
89.75
91.289
Nov 1, 2024
Feb 8, 2025
May 18, 2025
Aug 25, 2025
Dec 2, 2025
Mar 11, 2026
Jun 18, 2026
Average Accuracy
Worst-group Accuracy
ECE
ECE+TS
Updated 1mo ago
Evaluation Results
Method
Method
Links
Average Accuracy
Worst-group Accuracy
ECE
ECE+TS
ERM
2024.11
92.6
57.4
-
-
JTT
2024.11
92.6
63.7
-
-
Robust Filtered
2026.06
91.5
-
2.7
2.1
Robust MoE
2026.06
91.4
-
2.5
1.8
Single Expert
2026.06
91.3
-
6.9
6.4
Vanilla MoE
2026.06
91.2
-
7.1
6.5
MoCaE
2026.06
91.2
-
6.6
5.9
FGR
2026.06
90.6
-
4.4
3.5
FGR + Robust
2026.06
90.3
-
5.4
4.7
Bias-Conditioned Self-Influence
base_method=JTT
2024.11
86.9
78.5
-
-
Feedback
Search any
task
Search any
task