Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Cyber Red Teaming on CAGE 4 (200 episodes)
Loading...
9.86
AHD
Foundation-Sec + RL
9.854
9.8945
9.935
9.9755
May 16, 2026
AHD
ECR
AAC
Updated 2mo ago
Evaluation Results
Method
Method
Links
AHD
ECR
AAC
Foundation-Sec + RL
LLM Backbone=Foundatio...
2026.05
9.86
94
2,322
Foundation-Sec + RL + Reflexion
LLM Backbone=Foundatio...
2026.05
9.97
91
3,765
Qwen3 + RL
LLM Backbone=Qwen3, Re...
2026.05
10.01
100
3,559
Qwen3 + RL + Reflexion
LLM Backbone=Qwen3, Re...
2026.05
10.01
99.5
541
Feedback
Search any
task
Search any
task