Aviral Gupta

Hey, thanks for stopping by!

I’m a junior undergraduate in the Computer Science Department at BITS Pilani, deeply passionate about deep learning :). My research spans various areas in AI and DL, including diffusion models, fairness, and interpretability.

I am currently a research intern at the Vision and AI Lab, IISc Bangalore, where I work on ML fairness, bias mitigation, and domain generalization using synthetic data-driven approaches. My current focus is on unsupervised debiasing techniques through counterfactual dataset generation with diffusion models. Independently, my recent work and publications explore how interpretability methods can be enhanced using formal languages and structured representations.

More recently, I have been interested in multi-modal intepretability and mechanistic interpretability in vision. I am exploring principled methods to peer inside vision models leveraging MI techniques to understand and mitigate bias through interpretability-driven solutions.

This summer, I will also be interning at Trexquant as a ML Quant Researcher, where I will work on developing alphas for trading equities and futures markets using ML methods.

Always looking to collaborate on interesting problems, drop a mail!