Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Backdoor Attack on Fear Trigger Emotion (OOD)
Loading...
81.3
ASR (OOD)
GREAT
43.548
53.349
63.15
72.951
Oct 10, 2025
ASR (OOD)
Updated 1d ago
Evaluation Results
Method
Method
Links
ASR (OOD)
GREAT
Model=Llama-3.2-1B, Po...
2025.10
81.3
GREAT
Model=Llama-3.2-3B, Po...
2025.10
77.2
GREAT
Model=Llama-3.2-3B, Po...
2025.10
76.3
GREAT
Model=Llama-3.2-3B, Po...
2025.10
62.2
Random
Model=Llama-3.2-3B, Po...
2025.10
58.3
Random
Model=Llama-3.2-1B, Po...
2025.10
57.2
GREAT
Model=Llama-3.2-1B, Po...
2025.10
57
Random
Model=Llama-3.2-3B, Po...
2025.10
56.7
GREAT
Model=Llama-3.2-1B, Po...
2025.10
56.2
GREAT
Model=Llama-3.2-1B, Po...
2025.10
54
Random
Model=Llama-3.2-1B, Po...
2025.10
53.5
Random
Model=Llama-3.2-3B, Po...
2025.10
51.2
Random
Model=Llama-3.2-1B, Po...
2025.10
50.8
SUDO
Model=Llama-3.2-3B, Po...
2025.10
49.1
GREAT
Model=Llama-3.2-3B, Po...
2025.10
48.5
SUDO
Model=Llama-3.2-1B, Po...
2025.10
48.2
Random
Model=Llama-3.2-3B, Po...
2025.10
48
SUDO
Model=Llama-3.2-1B, Po...
2025.10
47
Random
Model=Llama-3.2-1B, Po...
2025.10
46.5
SUDO
Model=Llama-3.2-3B, Po...
2025.10
45
Feedback
Search any
task
Search any
task