Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Malicious activity detection on AgentDojo
Loading...
92.7
Precision
ADR
62.644
70.447
78.25
86.053
May 17, 2026
Precision
Recall
F1 Score
True Positives (TP)
False Positives (FP)
False Positive Rate (FPR)
Cost per Task ($)
Latency (s)
Cost per True Positive ($)
P95 Latency (s)
Updated 15d ago
Evaluation Results
Method
Method
Links
Precision
Recall
F1 Score
True Positives (TP)
False Positives (FP)
False Positive Rate (FPR)
Cost per Task ($)
Latency (s)
Cost per True Positive ($)
P95 Latency (s)
ADR
2026.05
92.7
100
96.2
38
3
5.5
0.012
18.5
0.029
47.3
ALRPHFS
2026.05
91.4
84.2
87.7
32
3
5.5
0.004
13.6
0.011
17
GuardAgent
2026.05
77.1
71.1
74
27
8
14.5
0.01
10.1
0.033
13.9
LlamaFirewall
2026.05
63.8
97.4
77.1
37
21
38.2
0.012
11.4
0.03
23
Feedback
Search any
task
Search any
task