Peer-Review
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Peer-Review Claude generated
64.02TPR@FPR=0.1%
14
Peer-Review Gemini generated
43.96TPR @ FPR=0.1%
14
Peer-Review GPT-4o generated
33.04TPR @ 0.1% FPR
14
Peer-review (evaluation set)
63RoBERTa ASR
12
Peer-Review
67.36TPR@0.1%
6
Peer-Review Claude
99.4AUROC
5
Peer-Review Gemini
97.9AUROC
5
Peer-Review GPT-4o
0.973AUROC
5