DiaASQ
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
DiaASQ
35LLM-as-a-Judge Score
10
DiaASQ EN (test)
0.479Pair F1 (Target-Aspect)
7
DiaASQ ZH (Chinese) (test)
48.61Pair Extraction F1 (Target-Aspect)
7
DiaASQ (test)
50LLM-as-a-Judge Score
5