Masked language model as a defense against textual adversarial attacks

A non-transitory storage medium having stored therein instructions that are executable by one or more hardware processors to perform operations comprising:

• receiving a text string that includes multiple words;
• tokenizing the text string to create a tokenized text string;
• substituting each token in the tokenized text string with a mask token to create a masked text string;
• performing an inference process on the masked text string to obtain a respective probability for each token;
• determining a respective suspicion level for each probability;
• modulating the suspicion levels to obtain a respective weighted suspicion score for each token;
• and
• comparing each of the weighted suspicion scores with a threshold to determine whether any one or more of the words indicate that the text string comprises an attack prompt.