Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Sentiment Analysis on SST2 POLISHED Attack (test)
Loading...
99
ASR
Unpatched
-2.92
23.54
50
76.46
Jun 2, 2026
ASR
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
ASR
Accuracy
Unpatched
Model=Llama
2026.06
99
95
FinePruning
Model=Llama
2026.06
98
94
FT_l2
Model=Llama
2026.06
97
94
ROME
Model=Llama
2026.06
97
92
FT_linf
Model=Llama
2026.06
96
94
SPP
Model=Llama
2026.06
96
91
FT
Model=Llama
2026.06
95
95
BAERASER
Model=Llama
2026.06
95
90
OneShot
Model=Llama
2026.06
95
90
MEND
Model=Llama
2026.06
94
93
Unpatched
Backbone=Qwen2.5-32B
2026.06
92
94
Mudjacking
Model=Llama
2026.06
82
90
Patcher
Model=Llama
2026.06
1
92
Patcher
Backbone=Qwen2.5-32B
2026.06
1
93
Feedback
Search any
task
Search any
task