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

Kk

Benchmarks

Task NameDataset NameSOTA ResultTrend
Logical ReasoningKK
Test Accuracy47.5
28
Logical ReasoningKK All
Accuracy53.6
21
Logical ReasoningKK 6–8 ppl.
Accuracy27.5
21
Logical ReasoningKK 4–5 ppl.
Accuracy60.3
21
Logical ReasoningKK 2–3 ppl.
Accuracy86.8
21
Logical ReasoningKK hard
Accuracy73.5
8
Logical ReasoningKK easy
Accuracy90
8
Machine Reading ComprehensionKk Kazakh
ROUGE-L6.9
4
Puzzle SolvingKK
Pass@1 Accuracy50.14
2
Showing 9 of 9 rows