Reproduce “Entity-based Reinforcement Learning for Autonomous Cyber Defence” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Isaac Symes Thompson, Alberto Caron, Chris Hicks, Vasilios Mavroudis
The authors present an entity-based Transformer policy for cyber-defence, representing network nodes as a variable-length collection rather than fixed positions in an input vector. It selects a node and defensive action from vulnerability and compromise records. On random 20-node networks, training on random 10-node networks yields mean reward 96.278664, close to 96.232763 for native 20-node training, in the tested simulator.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.