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

HQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Question AnsweringHQA
EM0.442
55
Knowledge gap detectionHQA
Accuracy81.5
18
Knowledge-Intensive ReasoningHQA
Average Score87
18
Question AnsweringHQA (val)
EM35.2
14
Question AnsweringHQA (in-domain)
EM39.6
14
Information RetrievalHQA (test)
Recall@557.7
7
Context Compression & QAHQA (val)
EM30.4
6
Showing 7 of 7 rows