Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Semantic Textual Similarity on STS-B (Accuracy/Bias evaluation)
Loading...
0.595
Accuracy
EAR
0.4546
0.49105
0.5275
0.56395
Jun 6, 2024
Accuracy
Bias
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Bias
EAR
Backbone=RoBERTa-base
2024.06
0.595
0.333
MABEL
Backbone=RoBERTa-base
2024.06
0.591
0.304
Vanilla-tuning
Backbone=RoBERTa-base
2024.06
0.578
0.33
MABEL
Backbone=BERT-base
2024.06
0.57
0.181
Debiased-tuning
Backbone=RoBERTa-base
2024.06
0.518
0.314
EAR
Backbone=BERT-base
2024.06
0.509
0.233
Vanilla-tuning
Backbone=BERT-base
2024.06
0.507
0.197
ProSocialTuning
Backbone=RoBERTa-base
2024.06
0.494
0.28
Debiased-tuning
Backbone=BERT-base
2024.06
0.473
0.184
ProSocialTuning
Backbone=BERT-base
2024.06
0.46
0.169
Feedback
Search any
task
Search any
task