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

Adversarial Attack on Mantis Eval

84.57Attack Success Rate

LAMP

28.607643.136357.66572.1937Jan 29, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
84.57
2026.01
82.71
2026.01
79.65
2026.01
75.44
2026.01
75.21
2026.01
73.43
2026.01
73.21
2026.01
66.68
2026.01
65.45
2026.01
64.68
2026.01
63.84
2026.01
60.79
2026.01
59.76
2026.01
57.58
2026.01
56.87
2026.01
56.74
2026.01
55.55
2026.01
54.79
2026.01
54.68
2026.01
54.68
2026.01
54.22
2026.01
53.68
2026.01
52.82
2026.01
52.74
2026.01
52.68
2026.01
52.13
2026.01
51.79
2026.01
51.66
2026.01
51.43
2026.01
50.74
2026.01
49.42
2026.01
48.41
2026.01
40.81
2026.01
36.15
2026.01
36.15
2026.01
33.15
2026.01
30.76