Traditionally, in Reinforcement Learning, the specification of the task Ls contained in the reinforcement function (RF), and ach new task requires the definition of a new RF. But in the nature, explicit reward signals are limited, and the characteristics of the environment afTects not only how animals ...