Search papers, labs, and topics across Lattice.
This study introduces Trident, a novel agentic LLM red teaming framework designed to evaluate the robustness of Deep Reinforcement Learning (DRL) cyber defenses against adaptive threats. By integrating a dynamic benchmark environment and a dataset of over 13,000 interaction trajectories, Trident reformulates red agent training into a contextual bandit problem, enabling the generation of sophisticated attack strategies. Empirical results demonstrate that Trident significantly outperforms static red agent baselines, revealing critical vulnerabilities in existing DRL defenses and showcasing emergent behaviors that traditional heuristics cannot detect.
Trident exposes a staggering 522% drop in defensive performance of DRL systems against adaptive threats, highlighting their critical vulnerabilities.
Autonomous cyber defense systems based on Deep Reinforcement Learning (DRL) have attracted significant research attention, yet remain evaluated almost exclusively against static, heuristic red agents, leaving their robustness against adaptive threats critically understudied. Meanwhile, recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) have improved LLM reasoning, but their integration into cybersecurity remains elusive due to the absence of suitable benchmark environments and interaction datasets. To bridge this gap, we introduce Trident, an agentic LLM red teaming framework comprising three components: a dynamic benchmark with isolated sandbox servers spanning CybORG CAGE 4 and CyberWheel, a dataset comprises over 13,000 high-fidelity red-blue interaction trajectories for RLVR, and a ``Code-as-Policy''RLVR agentic architecture Trident Agentic). The latter reformulates red agent training as a contextual bandit via a tripartite Log Summarizer--Planner--Coder design, where a trainable Planner generates complete attack strategies from compressed execution logs, which a frozen Coder translates into executable Python policies deployed against live DRL defenders. Empirical evaluations reveal a fundamental brittleness in existing defenses: with a single trainable 7B planner, Trident reduces blue agent defensive performance by an average of 522% compared to static red agent baselines while autonomously discovering emergent behaviors such as decoy avoidance and adaptive state prioritization that static heuristics entirely fail to uncover.
