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

WebQ

Benchmarks

Task NameDataset NameSOTA ResultTrend
CalibrationWebQ
ECE0.0674
31
Question AnsweringWebQ
EM32
27
Speech-to-Text Question-AnsweringWebQ
Accuracy66.6
23
Factuality EvaluationWebQ
Accuracy (Response)81
18
Speech-to-Speech Question-AnsweringWebQ
Accuracy61.5
13
Question AnsweringWebQ
Accuracy (WebQ)67.94
8
Multi-hop Question AnsweringWebQ 2013 (test)
F1 Score48.3
8
Question AnsweringWebQ
Exact Match37.76
7
Information RetrievalWebQ (test)
Top-20 Acc0.667
4
RetrievalWebQ
Accuracy (%)81.5
2
Showing 10 of 10 rows