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

Selective Generation on GSM8k

88.5ROC-AUC

Sheeps

38.99651.84864.777.552Aug 20, 2024
Updated 1mo ago

Evaluation Results

MethodLinks
2024.08
88.5
2024.08
88.2
2024.08
88.1
2024.08
85.3
2024.08
84.4
2024.08
83.9
2024.08
82.1
2024.08
81.9
2024.08
81.7
2024.08
81.6
2024.08
81.2
2024.08
80.9
2024.08
79.5
2024.08
79.5
2024.08
78.7
2024.08
78
2024.08
77.3
2024.08
76.1
2024.08
76
2024.08
75.6
2024.08
75.5
2024.08
74.6
2024.08
74.3
2024.08
73.9
2024.08
73.3
2024.08
73.1
2024.08
73
2024.08
72.6
2024.08
72.2
2024.08
72.1
2024.08
71.3
2024.08
70.3
2024.08
70.2
2024.08
69.9
2024.08
69.3
2024.08
69.1
2024.08
68.9
2024.08
68.5
2024.08
68.2
2024.08
68.1
2024.08
67.9
2024.08
67.8
2024.08
67.7
2024.08
67.6
2024.08
67.3
2024.08
67.2
2024.08
67.1
2024.08
67
2024.08
66.9
2024.08
66.8
2024.08
66.4
2024.08
66.4
2024.08
66.2
2024.08
66.1
2024.08
65.7
2024.08
65.5
2024.08
65.2
2024.08
63.7
2024.08
63.4
2024.08
63.4
2024.08
62.7
2024.08
61.4
2024.08
60.5
2024.08
55.8
2024.08
52.3
2024.08
40.9