Extremely thankful to the excellent guidance and help of my collaborators @SwapnilDreams10, @YamanKSingla, @KbalajiTweets, @cchangyou, @RatnRajiv at @midasIIITD @nyuniversity @Adobe @UBuffalo! As an undergraduate researcher this has been an amazing and a humbling experience.
Show more
Recent work in adversarial NLP has shown impressive results in trigger-based attacks, however these approaches are data intensive. MINIMAL a novel data-free approach mines input-agnostic adversarial triggers from models using the impressions produced with our data-free algorithm.
Show more
No need for data to craft adversaries anymore! 🧐 Check our paper "MINIMAL: Mining Models for Data Free Universal Adversarial Triggers" @RealAAAI #AAAI2022#. Extremely grateful for the 4 accept reviews as this will be my first A* publication in #NLProc#!
Show more
