Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful prompt detection on WJB
Loading...
97.55
F1 Score
MLPM
58.4356
68.5903
78.745
88.8997
Feb 22, 2025
F1 Score
Updated 1d ago
Evaluation Results
Method
Method
Links
F1 Score
MLPM
Backbone=OLMo2-7B-Inst...
2025.02
97.55
WildGuard
Methodology=Guard Model
2025.02
97.1
Ayub & Majumdar
Backbone=OLMo2-7B-Inst...
2025.02
96.84
GraniteGuardian-3-1-8B
Methodology=Guard Model
2025.02
96.75
MLPM
Backbone=Llama-8B-Inst...
2025.02
94.69
MLPM
Backbone=Mistral-7B-In...
2025.02
93.65
Abdelnabi et al.
Backbone=OLMo2-7B-Inst...
2025.02
93.48
Abdelnabi et al.
Backbone=Llama-8B-Inst...
2025.02
93.27
MLPM
Backbone=Qwen3-8B-Inst...
2025.02
92
Abdelnabi et al.
Backbone=Mistral-7B-In...
2025.02
91.13
Abdelnabi et al.
Backbone=Qwen3-8B-Inst...
2025.02
90.44
Ayub & Majumdar
Backbone=Mistral-7B-In...
2025.02
87.53
Ayub & Majumdar
Backbone=Llama-8B-Inst...
2025.02
82.85
Ayub & Majumdar
Backbone=Qwen3-8B-Inst...
2025.02
81.31
Aegis-Guard-D
Methodology=Guard Model
2025.02
75.44
LlamaGuard3
Methodology=Guard Model
2025.02
67.83
ShieldGemma-9B
Methodology=Guard Model
2025.02
59.94
Feedback
Search any
task
Search any
task