CAGE Challenge 4

CAGE Challenge 4 is a benchmark dataset used to evaluate multi-agent reinforcement learning methods in the context of network security response, focusing on operational constraints such as Mean Time to Recover (MTTR), false-positive response, and firewall change-management disruption.

Papers using CAGE Challenge 4 (2)

Safety-Contract Graph Multi-Agent Reinforcement Learning for Autonomous Network Security Response2026Interpreting Agent Behaviors in Reinforcement-Learning-Based Cyber-Battle Simulation Platforms2025