Rajeev Verma

Master's Student @ University of Amsterdam

“If we knew what it was we were doing, it would not be called research, would it?”— Albert Einstein

Hello! I am currently a second year master’s student supervised by Eric Nalisnick. Previously, I studied Electrical Engineering at the Indian Institute of Technology Patna (IITP). I was also affiliated with the AI-NLP-ML lab while at IITP.

My general research interests are in AI Safety and Responsible AI, and aim to build systems that are transparent, reliable, trustworthy, etc. To this end, I am excited about uncertainty quantification, Human-in-the-loop and Interactive Machine Learning, etc. In my free time, I also take keen interest in the state of peer-review system, and research on improving the quality of peer-review process by addressing problems like biasedness, arbitrariness, inconsistency, etc. I also organise a Statistics Reading Group.


Jun 13, 2022 New paper: On the Calibration of Learning to Defer to Multiple Experts accepted to the ICML’22 HMCaT Workshop. Work done with Daniel Barrejón and Eric Nalisnick.
May 15, 2022 Calibrated Learning to Defer with One-vs-All Classifiers will be presented at ICML’22.
Feb 8, 2022 New paper on the Calibration of Learning to Defer (L2D) systems: Calibrated Learning to Defer with One-vs-All Classifiers. We propose a new surrogate loss for L2D that is provably consistent and provides well-calibrated confidence estimates.
Nov 1, 2021 Started as ML/NLP researcher at CACTUS Communications.
Apr 1, 2021 Our MLRC reproducibility report was accepted to appear in ReScience Journal. Check it here.

selected publications

  1. ACL
    Deepsentipeer: Harnessing sentiment in review texts to recommend peer review decisions
    Ghosal, Tirthankar, Verma, Rajeev, Ekbal, Asif, and Bhattacharyya, Pushpak
    In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics 2019
  2. ICML
    Calibrated Learning to Defer with One-vs-All Classifiers
    Verma, Rajeev, and Nalisnick, Eric
    In Proceedings of the 39th International Conference on Machine Learning (ICML) 2022