MARL Adversarial

Policy Distillation

Knowledge transfer technique where a complex policy learned by an agent is compressed into a simpler and more efficient form, often used after adversarial training.

← Tillbaka