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

Safe-or-harmful binary classification on BeaverTails

84.6Accuracy

TELLME

51.52860.11468.777.286Feb 7, 2025
Updated 6d ago

Evaluation Results

MethodLinks
2025.02
84.6
2025.02
84.6
2025.02
84.5
2025.02
84.4
2025.02
84.4
2025.02
84.2
2025.02
83.8
2025.02
83.1
2025.02
82.9
2025.02
82.7
2025.02
82.5
2025.02
82.5
2025.02
82.5
2025.02
82.5
2025.02
82.3
2025.02
82.3
2025.02
82.1
2025.02
82.1
2025.02
82
2025.02
81.6
2025.02
81.5
2025.02
81.3
2025.02
81.2
2025.02
76.9
2025.02
75.6
2025.02
75.5
2025.02
75.4
2025.02
75.3
2025.02
75.2
2025.02
75.2
2025.02
75.2
2025.02
75.1
2025.02
74.8
2025.02
74.7
2025.02
74.7
2025.02
74.6
2025.02
73.8
2025.02
73.7
2025.02
73.4
2025.02
73.3
2025.02
73.1
2025.02
73.1
2025.02
73
2025.02
72.9
2025.02
72.8
2025.02
72.7
2025.02
72.1
2025.02
72
2025.02
71.9
2025.02
71.8
2025.02
71.8
2025.02
71.8
2025.02
71.6
2025.02
71.6
2025.02
70.4
2025.02
70.4
2025.02
65
2025.02
63.7
2025.02
55.6
2025.02
55.2
2025.02
53.6
2025.02
53.4
2025.02
52.8