Trident : How to Break Deep Reinforcement Learning Cyber Defenses (Agentic)

Ryozo Masukawa, Ian Bryant, Armita Kazeminajafabadi +6

Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, h…