CRUXEval
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
CRUXEval-O
83.5Accuracy
61
CRUXEval
98.8Input-CoT Accuracy
56
CRUXEval-O
79.1Pass@1
52
CRUXEval-I
67.5Pass@1
52
CRUXEval
76.87Accuracy
36
CRUXEval-O 200-sample (seed 42) (test)
55Accuracy
20
CRUXEval-O
87.5Score
19
CruxEval-O
62.7Pass@1
16
CruxEval-I
61.4Pass@1
16
CruxEval Output
51Score
12
CruxEval-O
36.4CX-O Score
10
CruxEval-I
35.7CX-I
10
CRUXEval I
74Accuracy
9
CRUXEval-O
41Pass@1
8
CRUXEval
88.5Pass@1
8
CruxEval
94.13Pass@1
6
CRUXEval L2
77Accuracy
4
CruxEval o
35.3Exact Match
4
CruxEval-i
36.2Exact Match
4
CruxEval
66.5Pass@1 (inv_step_call)
3
CRUXEval-O 1-shot
84.01Accuracy
3
CRUXEval-I 1-shot
79.75Accuracy
3
CRUXEval (L3)
74Accuracy
2
CRUXEval L1
89Accuracy
2