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

Explain

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hallucination predictionExplain (+ domain)
Accuracy64.87
20
Hallucination predictionExplain original
Accuracy80.91
20
Hallucination predictionExplain domain refined
AUROC70.04
10
Hallucination predictionExplain unrefined (original)
AUROC85.42
10
Showing 4 of 4 rows