CybORG++: An Enhanced Gym for the Development of Autonomous Cyber Agents
This provides a more efficient and customizable platform for developing autonomous cyber defence agents, benefiting researchers in enterprise network security, though it is incremental as it builds on an existing environment.
The researchers tackled the need for a better toolkit for reinforcement learning in network defence by developing CybORG++, which builds on CAGE 2 CybORG with improvements like enhanced debugging and a lightweight MiniCAGE version that achieves up to 1000x faster execution without losing accuracy.
CybORG++ is an advanced toolkit for reinforcement learning research focused on network defence. Building on the CAGE 2 CybORG environment, it introduces key improvements, including enhanced debugging capabilities, refined agent implementation support, and a streamlined environment that enables faster training and easier customisation. Along with addressing several software bugs from its predecessor, CybORG++ introduces MiniCAGE, a lightweight version of CAGE 2, which improves performance dramatically, up to 1000x faster execution in parallel iterations, without sacrificing accuracy or core functionality. CybORG++ serves as a robust platform for developing and evaluating defensive agents, making it a valuable resource for advancing enterprise network defence research.