Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Mental Manipulation Detection on PKU-SafeRLHF
Loading...
80
Accuracy
Dual-agent
60.24
65.37
70.5
75.63
Dec 1, 2025
Accuracy
Precision
Recall
F1 Score
ROC AUC
Spearman Rho
Pearson R
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy
Precision
Recall
F1 Score
ROC AUC
Spearman Rho
Pearson R
Dual-agent
Backbone=DeepSeek-R1
2025.12
80
73.4
94
82.5
84.3
0.665
0.66
MV (N = 40)
Backbone=Qwen-Plus
2025.12
72
64.1
100
78.1
81
0.639
0.639
Rule-based
2025.12
61
100
14.3
25
57.1
0.289
0.289
Feedback
Search any
task
Search any
task