Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

FRANK

Benchmarks

Task NameDataset NameSOTA ResultTrend
Factual Consistency EvaluationFRANK CNNDM
Spearman Correlation61.8
30
Factual Consistency EvaluationFRANK-XSum (FRK-X)
Spearman Correlation32.1
30
Factual Consistency EvaluationFRANK CNNDM (test)
PCC67.7
22
Factual Consistency EvaluationFRANK-XSum (test)
Pearson Correlation Coefficient38.3
22
Hallucination DetectionFRANK
Balanced Acc77.2
18
Faithfulness evaluationFRANK
Pearson Corr0.841
10
Factual Consistency EvaluationFRANK CNNDM
Pearson R68.9
8
Factual Consistency EvaluationFRANK XSum
Pearson Correlation Coefficient38.9
8
Factuality Error LocalizationFRANK
Accuracy (OutE)56.7
3
Showing 9 of 9 rows