Biography

I am a first year Computer Science PhD student at the USC Viterbi School of Engineering. Previously, I was a Research Fellow with the Information Extraction and Synthesis Laboratory (IESL) at the University of Massachusetts (UMass), Amherst.

I am interested in research on natural language processing (NLP) systems that can reason over text and knowledge bases (KBs). I have worked on retrieval for open-domain question answering (ODQA), knowledge base completion (KBC), and semantic parsing for knowledge base QA (KBQA).

I sporadically blog (I mostly edit and proof-read) on Medium.

Interests
  • Artificial Intelligence
  • Machine Learning
  • Natural Language Processing
  • Reinforcement Learning
Education
  • PhD in Computer Science, 2021-Present

    University of Southern California

  • MSc in Computer Science, 2020

    University of Massachusetts, Amherst

  • BTech in Electronics and Communication Engineering, 2018

    Indian Institute of Technology (IIT), Guwahati

Experience

 
 
 
 
 
Research Fellow
Jun 2020 – Jun 2021 Massachusetts

The role allowed me to pursue projects with the IESL lab members. In addition, I contribute to OpenReview, the open-access, open-discussion conference platform. I also work with the Chan Zuckerberg Initiative to apply IESL’s lab output to CZI’s systems and databases.

Tasks:

  • Case-based reasoning for knowledge base completion and question answering.
  • Incorporated a language model based system trained on citation and authorship graphs into OpenReview Expertise which generates affinity scores between submitted papers and available reviewers. This system generates affinity scores between submitted papers and available reviewers' past body of work.
  • Added features to the fairness-constrained matching algorithms of OpenReview Matcher, which solve an optimization problem to assign papers for review given pre-computed affinity scores. The “fair” matching algorithms ensure that the reviewers assigned to every submission possess a (combined) minimum level of familiarity with the research area.
  • Worked on scaling an affiliation disambiguation system to 400K+ target affiliation labels as part of IESL’s collaboration with the Meta-CZI.
 
 
 
 
 
Machine Learning Intern
May 2019 – Aug 2019 California

Member of a team of researchers from the Artificial Intelligence Center (AIC), SRI International participating in the DARPA program: Radio Frequency Machine Learning Systems (RFMLS)

Tasks:

  • Developed and maintained an RF spectrum simulator for fast and reproducible experimentation.
  • Implemented and benchmarked reinforcement learning approaches to control/steer an antenna array for RF monitoring.
  • Experimented with alternative training regimes such as imitation learning. This method led to better performance than the available baselines.
 
 
 
 
 
Software Development Intern
May 2016 – Jul 2016 Pune, India
  • Studied and implemented the principles of parallel processing using MPI and OpenMP.
  • Contributed to a molecular dynamics simulator capable of leveraging the compute capabilities of a CPU cluster. The simulator implemented distributed computation of the physics of particle movement.