The team introduced a novel Hybrid-WM reinforcement learning model, which incorporates working memory limitations into the decision-making framework. "Unlike older models, this new framework takes ...