I am now an AI Research Engineer at Aleph Alpha in Germany where I work on all things LLM inference. We build infra to efficiently serve LLMs while also allowing enhancements at inference time including steering and explainability. For an overview of some of my work at Aleph Alpha, check out this short(-ish) talk.
I got my PhD in Computer Science from University of Maryland, College Park. I was part of the Maryland-Max Planck joint program through which I also spent time at Max Planck Institute for Software Systems (MPI-SWS). Throughout my PhD, I was fortunate to have the guidance of Krishna P. Gummadi (MPI-SWS) and
John P. Dickerson (University of Maryland) as my advisors, and collaborated closely with Adrian Weller at the University of Cambridge.
My research focuses on building a better understanding of trustworthy deep learning through an empirical lens.
I've contributed to research in fairness (ICML2019 AAAI2020 AAAI2023),
robustness (ICML2022 & AAAI2023), and my work was one of the first to establish a connection between the two fields (FAccT2021).
Lately, in the era of large models, I have been thinking about efficient learning (NeurIPS2023) and taught a seminar on Systems for LLMs.
During my PhD, I had the pleasure of interning twice at Amazon where I first worked on counterfactual explanations with AWS Clarify and then on fairness aspects of generative AI with AWS Bedrock.
I obtained my undergrad from IIIT Delhi (2015 - 2019) majoring in Computer Science and Engineering, where I was associated to Precog. During my undergrad I worked on topics related to Social Computing, Computational Social Science and ICT4D.
I am fortunate to have worked with some amazing mentors (in alphabetical order): Soheil Feizi (UMD), Tom Goldstein (UMD), Hoda Heidari (ETH Zürich), Ponnurangam Kumaraguru (IIIT Delhi), Bradley C. Love (UCL), Rijurekha Sen (IIT Delhi), Pushpendra Singh (IIIT Delhi), Mariya Toneva (MPI-SWS), and Muhammad Bilal Zafar (Amazon). I spent the wonderful summer of 2018 interning at MPI-SWS where I was advised by Krishna P. Gummadi.
For more, check out my CV.
vnanda [at] mpi-sws [dot] org