Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety Evaluation on Anchor Safety Dataset
Loading...
100
Anchor Accuracy
ORIGINAL
-4
23
50
77
Mar 19, 2026
Anchor Accuracy
Updated 29d ago
Evaluation Results
Method
Method
Links
Anchor Accuracy
ORIGINAL
Model Family=GEMMA-2-2B
2026.03
100
CDA
Model Family=GEMMA-2-2B
2026.03
100
UGID
Model Family=GEMMA-2-2B
2026.03
100
ORIGINAL
Model Family=QWEN2.5-3B
2026.03
100
CDA
Model Family=QWEN2.5-3B
2026.03
100
UGID
Model Family=QWEN2.5-3B
2026.03
100
ORIGINAL
Model Family=QWEN2.5-7B
2026.03
100
CDA
Model Family=QWEN2.5-7B
2026.03
100
UGID
Model Family=QWEN2.5-7B
2026.03
100
ORIGINAL
Model Family=QWEN2.5-14B
2026.03
100
KLAAD
Model Family=QWEN2.5-14B
2026.03
100
UGID
Model Family=QWEN2.5-14B
2026.03
100
KLAAD
Model Family=GEMMA-2-2B
2026.03
50
KLAAD
Model Family=QWEN2.5-3B
2026.03
50
CDA
Model Family=QWEN2.5-14B
2026.03
50
KLAAD
Model Family=QWEN2.5-7B
2026.03
0
Feedback
Search any
task
Search any
task