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

Bug injection on MBPP+

99.89Attack Success Rate (ASR)

XOXO + GCGS

38.32254.30670.2986.274Mar 18, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.03
99.8920
2025.03
99.8822
2025.03
99.8822
2025.03
99.5525
2025.03
99.3637
2025.03
98.9943
2025.03
98.4275
2025.03
96.4180
2025.03
95.2784
2025.03
84.53169
2025.03
81.29180
2025.03
75.37235
2025.03
73.17228
2025.03
40.69233