Hello, I am an Associate Professor (Senior Lecturer) at the Computer Science department of the University of Bath, UK, and a visiting Professor at Aalto University, Finland. I work on large-scale machine learning problems particularly in Extreme Classification, sparse neural network training and learning with imperfect supervision. Looking for motivated research assistants, please email at firstname.lastname@aalto.fi
Before this, I was a post-doc at Max-Planck Institute for Intelligent Systems, Tuebingen, Germany in the group of Prof. Bernhard Schölkopf. I finished my PhD from University of Grenoble, France where I was advised by Prof. Eric Gaussier and Prof. Massih-Reza Amini, and a Master degree from Chennai Mathematical Institute.
I am looking for motivated PhD students to join my group at the University of Bath, UK. We are working on exciting problems on sparse and memory/energy efficient neural networks for large output spaces, and Large Language Models going forward. Two of our recent works in this area have been published in NeurIPS 2024 and ECML 2023. Please reach out at rb2608--AT--bath.ac.uk, if you are interested in joining!
Selected Publications (Google scholar)
Labels in Extremes: How Well Calibrated are Extreme Multi-label Classifiers? ( pdf , code )
Nasib Ullah, Erik Schultheis, Jinbin Zhang and Rohit Babbar
31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD, 2025, Toronto, CanadaLarge Language Model as a Teacher for Zero-shot Tagging at Extreme Scales ( pdf , code )
Jinbin Zhang, Nasib Ullah, and Rohit Babbar
31st International Conference on Computational Linguistics, COLING, 2025, Abu-Dhabi, UAENavigating Extremes : Dynamic Sparsity in Large Output Spaces ( pdf , code )
Nasib Ullah, Erik Schultheis, Mike Lasby, Yani Ioannou and Rohit Babbar
38th Conference on Neural Information Processing Systems, NeurIPS, 2024, Vancouver, CanadaGandalf : Learning Label-label correlations in Extreme Multi-label Classification via Label Features (pdf )
Siddhant Kharbanda, Devaansh Gupta, Erik Schultheis, Atmadeep Banerjee, Cho-Jui Hsieh, and Rohit Babbar
30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2024, Barcelona, SpainA General Online Algorithm for Optimizing Complex Performance Metrics
Erik Schultheis, Marek Wydmuch, Wojciech Kotlowski, Rohit Babbar, and Krzysztof Dembczyński
41st International Conference on Machine Learning, ICML 2024, Vienna, AustriaConsistent Algorithms for Multi-label Classification with Macro@k Metrics ( pdf , code )
Erik Schultheis, Marek Wydmuch, Wojciech Kotlowski, Rohit Babbar, Strom Borman, and Krzysztof Dembczyński
12th International Conference on Learning Representations, ICLR 2024, Vienna, AustriaGeneralized Test Utilities for Long-tail Performance in Extreme Multi-label Classification ( pdf, code )
Erik Schultheis, Marek Wydmuch, Wojciech Kotlowski, Rohit Babbar, and Krzysztof Dembczyński
37th Conference on Neural Information Processing Systems, NeurIPS, 2023, New Orleans, USATowards Memory-Efficient Training for Extremely Large Output Spaces -- Learning with 500k Labels on a Single Commodity GPU
Erik Schultheis, Rohit Babbar
ECML-PKDD 2023, (pre-print, code ), Torino, ItalyInceptionXML : A Lightweight Framework with Synchronized Negative Sampling for Short Text Extreme Classification (pdf , code )
Siddhant Kharbanda, Atmadeep Banerjee, Devaansh Gupta, Akash Palrecha, and Rohit Babbar
46th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2023, Taipei, TaiwanCascadeXML : Rethinking Transformers for End-to-end Multi-resolution Training in Extreme Multi-Label Classification (pdf , code),
Siddhant Kharbanda, Atmadeep Banerjee, Erik Schultheis and Rohit Babbar
36th Conference on Neural Information Processing Systems, NeurIPS, 2022, New Orleans, USAOn Missing Labels, Long tails and propensities in Extreme Multi-label Classification (pdf , code)
Erik Schultheis, Marek Wydmuch, Rohit Babbar, and Krzysztof Dembczyński
28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD 2022, Washington, USAAdversarial Examples for Extreme Multi-label Classification, (pdf, code)
Mohamaadreza Qaraei, Rohit Babbar, Machine Learning Journal 2022Speeding-up One-vs-All Training for Extreme Classification via Mean-separating Initialization, (pdf, Code)
Erik Schultheis, and Rohit Babbar
ECML & Machine Learning Journal Track, ECML-MLJ, 2022Convex Surrogates for Unbiased Loss Functions in Extreme Classification With Missing Labels (pdf, video, code)
Mohammadreza Qaraei, Erik Schultheis, Priyanshu Gupta, Rohit Babbar
30th ACM International World Wide Web Conference, The WebConf (WWW) 2021