Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Jailbreak evaluation on curated dataset (test)
Loading...
0
BAD BOT Rate
base64_input_only
-0.04
0.23
0.5
0.77
Jul 5, 2023
BAD BOT Rate
GOOD BOT Rate
UNCLEAR Rate
Updated 4d ago
Evaluation Results
Method
Method
Links
BAD BOT Rate
GOOD BOT Rate
UNCLEAR Rate
base64_input_only
Model=GPT-3.5 Turbo, E...
2023.07
0
0.53
0.47
base64_output_only
Model=GPT-3.5 Turbo, E...
2023.07
0
0.09
0.91
base64_raw
Model=GPT-3.5 Turbo, E...
2023.07
0
0
1
none
Model=GPT-3.5 Turbo, C...
2023.07
0.03
0.97
0
base64
Model=GPT-3.5 Turbo, E...
2023.07
0.03
0.06
0.91
style_injection_short
Model=GPT-3.5 Turbo
2023.07
0.69
0.19
0.12
dev_mode_v2
Model=GPT-3.5 Turbo, A...
2023.07
0.78
0.22
0
evil_system_prompt
Model=GPT-3.5 Turbo, A...
2023.07
0.88
0.09
0.03
AIM
Model=GPT-3.5 Turbo, A...
2023.07
0.97
0.03
0
dev_mode_with_rant
Model=GPT-3.5 Turbo, A...
2023.07
0.97
0.03
0
Adaptive attack
Model=GPT-3.5 Turbo
2023.07
1
0
-
Feedback
Search any
task
Search any
task