By calculating the Kullback-Leibler divergence between two probability measures belonging to different exponential families, we end up with a formula that generalizes the ordinary Fenchel-Young divergence which is recovered in the special case when we let the two exponential families coincide. Inspired by this formula, we define the duo Fenchel-Young divergence and reports a dominance condition on its pair of generators which guarantees that it is always non-negative.

翻译：通过计算属于不同指数家庭的两种概率计量方法之间的库尔贝克-利伯尔差异,我们最终得出了一个公式,将普通的Fenchel-Young差异概括化,在特例中,当我们让这两个指数型家庭重合时,在这两个指数型家庭重合时,就收回了这种差异。 受这个公式的启发,我们定义了deo Fenchel-Young差异,并报告了其发电机的主导条件,这保证了这些发电机始终是非负的。