Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety Alignment on HEx-PHI
Loading...
98.8
HEx-PHI Score
DiaBlo
77.896
83.323
88.75
94.177
Jun 3, 2025
Jul 30, 2025
Sep 25, 2025
Nov 21, 2025
Jan 17, 2026
Mar 15, 2026
May 11, 2026
HEx-PHI Score
Updated 14d ago
Evaluation Results
Method
Method
Links
HEx-PHI Score
DiaBlo
Model=Mistral-7B, r/N=...
2025.06
98.8
DiaBlo
Model=Mistral-7B, r/N=...
2025.06
98.8
DiaBlo
Model=Llama3-8B, r/N=N...
2025.06
97.6
HELLoRA
Params=0.31%, Backbone...
2026.05
96.67
DiaBlo
Model=Llama3-8B, r/N=N...
2025.06
96.3
HELLoRI-S
Params=0.02%, Backbone...
2026.05
96.06
DoRA
Model=Mistral-7B, r/N=...
2025.06
95.3
HELLoRI-D
Params=0.17%, Backbone...
2026.05
94.85
LoRI
Model=Mistral-7B, r/N=...
2025.06
94.7
LoRA
Params=1.03%, Backbone...
2026.05
93.94
LoRI-S
Params=0.06%, Backbone...
2026.05
93.64
DoRA
Model=Llama3-8B, r/N=r...
2025.06
93.6
LoRI-D
Params=0.59%, Backbone...
2026.05
93.33
LoRI
Model=Llama3-8B, r/N=r...
2025.06
92.8
LoRA
Model=Mistral-7B, r/N=...
2025.06
91.9
LoRA
Model=Llama3-8B, r/N=r...
2025.06
91.6
Zero-shot
Model=Mistral-7B, r/N=...
2025.06
80.5
Zero-shot
Model=Llama3-8B, r/N=N...
2025.06
78.7
Feedback
Search any
task
Search any
task