Analyzing fake-news detectors under adversarial threat using the
Text-Attack Library for a number of model hyper-parameters and attack
settings.