Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Safety Evaluation (FRR) on XSTest
Loading...
1.6
FRR
JPU
0.2072
9.6086
19.01
28.4114
Jan 6, 2026
FRR
Updated 4d ago
Evaluation Results
Method
Method
Links
FRR
JPU
Backbone=Llama3-8B-Ins...
2026.01
1.6
Base model
Backbone=Llama3-8B-Ins...
2026.01
3.2
CKU
Backbone=Llama3-8B-Ins...
2026.01
5.11
Eraser
Backbone=Llama3-8B-Ins...
2026.01
5.22
Safe Unlearning
Backbone=Llama3-8B-Ins...
2026.01
5.44
Circuit Break
Backbone=Llama3-8B-Ins...
2026.01
5.56
RSFT
Backbone=Llama3-8B-Ins...
2026.01
6
JPU
Backbone=Llama2-7B-Chat
2026.01
22.4
CKU
Backbone=Llama2-7B-Chat
2026.01
25.56
Safe Unlearning
Backbone=Llama2-7B-Chat
2026.01
27.78
Base model
Backbone=Llama2-7B-Chat
2026.01
28.8
Circuit Break
Backbone=Llama2-7B-Chat
2026.01
29.78
Eraser
Backbone=Llama2-7B-Chat
2026.01
33.33
RSFT
Backbone=Llama2-7B-Chat
2026.01
36.42
Feedback
Search any
task
Search any
task