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

Target behavior attack on MBPP

99.2ASR

AiTM

-3.96822.81649.676.384Feb 20, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.02
99.2
2025.02
98.5
2025.02
96.9
2025.02
96.3
2025.02
95.9
2025.02
95.1
2025.02
92.4
2025.02
92.3
2025.02
90.5
2025.02
80.4
2025.02
76.8
2025.02
73.1
2025.02
69.3
2025.02
55.9
2025.02
0
2025.02
0