Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Non-control benchmark families

Benchmarks

Task NameDataset NameSOTA ResultTrend
Deception Detection7 non-control benchmark families Mean across targets
AUROC91.6
3
Deception Detection7 non-control benchmark families GEMMA4-26B-A4B-IT
AUROC91.1
3
Deception Detection7 non-control benchmark families GPT-OSS-20B
AUROC0.92
3
Showing 3 of 3 rows