Dual-reinforcement learning theory proposes behaviour is under the tutelage of a retrospective, value-caching, model-free (MF) system and a prospective-planning, model-based (MB), system. This architecture raises a question ...