Tag
ASGARD proposes a two-phase teacher–student pipeline using reinforcement learning to enhance UAV resilience against action-space attacks, ensuring mission completion through corrected commands and generalization to unseen attacks.
This paper systematically investigates failure modes in reinforcement learning for small language models (70-500M parameters) using PPO, identifies silent LoRA freezing, numerical overflow, and catastrophic policy collapse, and proposes a robust system with merge-and-reinitialize adapters, float32 precision, and a safety mechanism. The approach converges stably and outperforms baselines with less data.