Abstract: Rationalization empowers deep learning models with self-explaining capabilities through a cooperative game, where a generator selects a semantically consistent subset of the input as a ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results