Harnessing the linguistic signal to predict scalar inferences

ACL 2020