Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
LLM Jailbreaking on AdaSteer Evaluation Set (test)
Loading...
1
SRF
SCAV
-1.56
15.72
33
50.28
May 19, 2026
SRF
HB
SR
Updated 13d ago
Evaluation Results
Method
Method
Links
SRF
HB
SR
SCAV
Target LLM=Llama-3.1-8...
2026.05
1
0
0
RepE
Target LLM=Gemma-2-9b-...
2026.05
1
0
1
Angular
Target LLM=Gemma-2-9b-...
2026.05
2
1
0
RepE
Target LLM=Qwen2.5-7B-...
2026.05
4
5
6
RepE
Target LLM=Llama-3.1-8...
2026.05
11
21
21
RD-C
Target LLM=Gemma-2-9b-...
2026.05
18
26
29
SCAV
Target LLM=Qwen2.5-7B-...
2026.05
23
51
37
SCAV
Target LLM=Gemma-2-9b-...
2026.05
30
47
53
RD-A
Target LLM=Gemma-2-9b-...
2026.05
33
35
42
Adaptive Probe-based Steering
Target LLM=Gemma-2-9b-...
2026.05
50
51
73
Angular
Target LLM=Llama-3.1-8...
2026.05
62
73
82
Adaptive Probe-based Steering
Target LLM=Qwen2.5-7B-...
2026.05
62
73
82
Adaptive Probe-based Steering
Target LLM=Llama-3.1-8...
2026.05
65
79
88
Angular
Target LLM=Qwen2.5-7B-...
2026.05
65
75
84
Feedback
Search any
task
Search any
task