E-SNLI
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
E-SNLI
91.31Accuracy
46
e-SNLI
89.6Accuracy
16
e-SNLI
78.44Accuracy
10
e-SNLI (test)
94Accuracy
9
e-SNLI
41Initial Performance
8
e-SNLI
50Human Evaluation Score
7
e-SNLI (out-domain)
2.98Grammar Score
7
e-SNLI abundant
88.8Accuracy
6
e-SNLI (medium)
87.5Accuracy
6
e-SNLI scarce
86.3Accuracy
6
e-SNLI (test)
86.66Accuracy
6
e-SNLI (test)
3.49GPT-4 Score
6
E-SNLI 10k queries (val)
92.4NDCG@5
4
e-SNLI
4.35ECE
4
e-SNLI 60-shot
40.1Accuracy
3