Kempner Institute for the Study of Natural and Artificial Intelligence, Harvard University, Cambridge, MA 02134, U.S.A.
Department of Psychology and Center for Brain Science, Harvard University, Cambridge, MA 02134, U.S.A.
Neural Comput. 2024 Oct 11;36(11):2225-2298. doi: 10.1162/neco_a_01705.
Adaptive behavior often requires predicting future events. The theory of reinforcement learning prescribes what kinds of predictive representations are useful and how to compute them. This review integrates these theoretical ideas with work on cognition and neuroscience. We pay special attention to the successor representation and its generalizations, which have been widely applied as both engineering tools and models of brain function. This convergence suggests that particular kinds of predictive representations may function as versatile building blocks of intelligence.
适应行为通常需要预测未来事件。强化学习理论规定了哪些预测表示是有用的，以及如何计算它们。本综述将这些理论思想与认知和神经科学的研究结合起来。我们特别关注后继表示及其推广，它们已被广泛应用于工程工具和大脑功能模型。这种趋同表明，某些特定类型的预测表示可能是智能的多功能构建块。
