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

Biography

Benchmarks

Task NameDataset NameSOTA ResultTrend
Long-form Question AnsweringBiography FullP (100% polluted)
VeriScore F1@k61.87
26
Long-form Question AnsweringBiography MixP (50% polluted)
VeriScore F1@k70.11
26
Long-form Question AnsweringBiography
VeriScore F149.7
14
Factuality EvaluationBiography
Correctness Count29.4
12
Long-form factuality evaluationBiography
Accuracy65.9
7
Showing 5 of 5 rows