Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Toxicity Steering on 15 prefix prompts length 50
Loading...
71.2
Toxicity Accuracy
ILRR
-2.016
16.992
36
55.008
Jan 29, 2026
Toxicity Accuracy
Updated 3d ago
Evaluation Results
Method
Method
Links
Toxicity Accuracy
ILRR
Base Model=LLaDA, alph...
2026.01
71.2
ILRR
Base Model=LLaDA, alph...
2026.01
60.7
ILRR
Base Model=MDLM, alpha...
2026.01
15.7
FK
Base Model=LLaDA, Sequ...
2026.01
9
PG-DLM
Base Model=LLaDA, Sequ...
2026.01
8.3
ILRR
Base Model=MDLM, alpha...
2026.01
7.2
FK
Base Model=MDLM, phi=1...
2026.01
3.8
best-of-n
Base Model=LLaDA, Sequ...
2026.01
2.4
best-of-n
Base Model=MDLM, Seque...
2026.01
1.9
PG-DLM
Base Model=MDLM, Seque...
2026.01
1.4
FK
Base Model=MDLM, phi=4...
2026.01
0.8
Feedback
Search any
task
Search any
task