Review
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Review (test)
95.76Accuracy
10
Review Interpolated Users
84.6Winrate
8
Review Trained Users
92.3Winrate
8
Review 1 821 abstracts (Final Includes)
45False Positives
8
Review 1
18False Positives
8
Review (test)
0.959AUROC
7
Review (test)
0.827AUROC
7
Review-5K
1.957MAE
5
Review 2 (7741 abstracts)
87False Positives (FP)
5
Review 2 (Final Includes)
—Primary metric
0