“Graded Suspiciousness of Adversarial Texts to Humans”. 2025. Computational Linguistics 51 (3). https://cljournal.org/index.php/cljournal/article/view/2771.