Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
LLM Robustness Evaluation on Robustness Diagnostics LLaMA-3-8B
Loading...
8.92
Template Mean
ORIGINAL
0.6624
2.8062
4.95
7.0938
Mar 19, 2026
Template Mean
Template Variance
Directional Gap
Neutral Mass
Updated 1mo ago
Evaluation Results
Method
Method
Links
Template Mean
Template Variance
Directional Gap
Neutral Mass
ORIGINAL
Backbone=LLaMA-3-8B
2026.03
8.92
3.41
1.87
0.02
KLAAD
Backbone=LLaMA-3-8B
2026.03
1.32
0.287
0.025
0
CDA
Backbone=LLaMA-3-8B
2026.03
1.21
0.0181
0.15
0
UGID
Backbone=LLaMA-3-8B
2026.03
0.98
0.0044
0.0625
0.0133
Feedback
Search any
task
Search any
task
