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

MultiHiertt

Benchmarks

Task NameDataset NameSOTA ResultTrend
Hierarchical Table Question AnsweringMultiHiertt (test)
Accuracy70.32
30
Table Question AnsweringMultiHiertt original (dev)
Exact-match Accuracy71.17
14
Question AnsweringMULTIHIERTT (test)
EM83.12
11
Tabular ReasoningMultiHierTT
Pass@1 Accuracy28.7
10
Program GenerationMultiHiertt
Program Accuracy82.9
10
Question AnsweringMULTIHIERTT (dev)
EM37.05
10
Multi-table Question AnsweringMultiHiertt
Exact Match (EM)22.99
7
Financial Table Question AnsweringMultiHiertt
Program Accuracy70.01
4
General Tabular ReasoningMultiHiertt (dev)
Accuracy71.17
3
Showing 9 of 9 rows