4 min readResearch note
Autonomous Cyberdefense Needs More Than an LLM
Autonomous cyberdefense with LLMs: control states, action boundaries, recovery and separate evaluation of utility and safety.
Research Notes on Federated Learning, Cybersecurity and Cyberdefense
Research notes and technical guides on federated learning, distributed AI, cyberdefense, trustworthy systems, and privacy-preserving platforms.
9 articles. Page 1 of 2.
4 min readResearch note
GraphRAG for Cyber Threat Intelligence
GraphRAG for threat intelligence: entities, relationships, STIX, provenance and evaluation against passage-based retrieval.
5 min readResearch note
How to Measure Robustness in Adversarial DFL Experiments
Evaluate DFL robustness with F1, false rejection, overhead and recovery. Interpret Flighter and compare clean and attacked conditions.
5 min readResearch note
Flighter: Situational Awareness as a Reliability Signal in DFL
How Flighter combines context and models to evaluate DFL reliability: study scope, indicator ablations and participant recovery.
5 min readResearch note
From Monitoring to Mitigation: A DFL Cyberdefense Lifecycle with LLM Explanations
Connect telemetry, federated learning, alerts and LLM-assisted mitigation with traceable evidence, validation and reviewed feedback.
3 min readResearch note
Situational Awareness for Cyberdefense with Decentralized Federated Learning
Build cyberdefense situational awareness with DFL: provenance, coverage, uncertainty and shared context for operational decisions.