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

ReClor

Benchmarks

Task NameDataset NameSOTA ResultTrend
Logical ReasoningReClor (test)
Accuracy80.6
87
Logical Reasoning Machine Reading ComprehensionReClor
Accuracy80
54
Logical ReasoningReClor (dev)
Accuracy0.786
46
Logical ReasoningReClor
Accuracy85.43
39
Logical ReasoningReClor Hard (test)
Accuracy87.2
37
Logical ReasoningReClor Easy (test)
Accuracy86.4
28
Logical ReasoningReClor (test-H)
Accuracy75
23
Logical ReasoningReClor (test-e)
Accuracy91.8
23
Logical ReasoningReClor v1 (test)
Accuracy78.1
23
Logical ReasoningReClor (val)
Accuracy (ReClor val)84.4
15
Logical ReasoningReClor 1.0 (test-H)
Accuracy72.9
13
Logical ReasoningReClor
Pass@161.8
10
Logical ReasoningReClor EASY v1
Accuracy81.4
9
Logical ReasoningReClor 1.0 (test-E)
Accuracy79.09
8
Logical ReasoningReClor 1.0 (val)
Accuracy68.4
8
Logical ReasoningReClor v1 (test-e)
Accuracy85.2
5
Logical ReasoningReClor v1 (dev)
Accuracy81.8
5
Question AnsweringReClor
Accuracy36.67
2
Logical ReasoningReClor 300 examples (val)
Accuracy2.7
1
Showing 19 of 19 rows