Brandon Laughlin, Christopher Collins, Karthik Sankaranarayanan +1
This paper presents a framework which enables a user to more easily make corrections to adversarial texts. While attack algorithms have been demonstrated to automatically build adv…