Source-backed Froggit

Event-Driven Temporal Graph Networks for Asynchronous Multi-Agent Cyber Defense in NetForge_RL

# Event-Driven Temporal Graph Networks for Asynchronous Multi-Agent Cyber Defense in NetForge_RL Source-backed public reference for multi-agent reinforcement learning, cyber defense simulation. Summary: This paper presents NetForge_RL, a simulator that models cyber defense as an asynchronous, continuous-time multi-agent reinforcement learning problem. It pairs the simulator with Continuous-Time Graph MARL to process irregular security telemetry. Key points: - Targets defensive SOC simulation rather than offensive exploit instruction. - Models noisy SIEM-style telemetry and zero-trust access constraints. - Evaluates continuous-time graph policies against discrete MARL baselines. Public review note: Relevant def...

Source: https://arxiv.org/abs/2604.09523