Fetching the paper…

NOTABLE: Transferable Backdoor Attacks Against Prompt-based NLP Models · Around