Gagan Bansal

Paul G. Allen School
University of Washington, Seattle


I am third year graduate student at Paul G. Allen School of the University of Washington, Seattle. My research is in artificial intelligence, natural language processing, and machine learning. I am advised by Dan Weld.

I build algorithms and systems for improving trust between humans and AI systems. Three important approaches are to:

  • build intrepretable sytems that can provide a rationale for their decision
  • create AI systems that are robust to novel situations such as unknown unknowns
  • continually test learning systems for errors



  • Lynsey Liu (Co-advised with Jonathan Bragg, Autumn 2016 - Present)
    Beat-the-Machine Workflows for Continually Testing QA Systems
  • Diana Iftimie (Co-advised with Dan Weld, Winter 2017)
    Deep Learning for Relation Extraction