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

Citations

Benchmarks

Task NameDataset NameSOTA ResultTrend
Citation Hallucination DetectionCitations PDF input mode rendered benchmark synthetic and real-world
Total Citations Count1,132
16
Citation Hallucination DetectionCitations BibTeX input mode synthetic and real-world (source .bib entries)
Total Count1,156
14
Text ClassificationCitations (full)
Accuracy64.6
8
CalibrationCitations (test)
ECE0.179
2
Showing 4 of 4 rows