I am a Research Fellow (CommBank AI Scholar) at Australian Institute for Machine Learning, Adelaide University working with Prof. Simon Lucey. I did my doctorate in Machine Learning theory advised by Prof. Debarghya Ghoshdastidar at the Theoretical Foundations of Artificial Intelligence group, Technical University of Munich.

My research focuses on the foundations of deep learning, with an emphasis on (i) understanding what deep neural networks learn, and (ii) characterizing the robustness of their predictions to adversarial corruptions of the data. Methodologically, I draw on statistical learning theory and kernel views of deep networks to obtain tractable analyses and transferable insights. I am also interested in learning on graphs.

Before my PhD, I did Masters of Science in Informatics at Technical University of Munich and bachelors in Computer Science & Engineering from National Institute of Technology, Trichy, India. Between my undergraduate and graduate studies, I spent a little over 3 years at Adobe Systems, India as a Computer Scientist.

Research Interests

  • Deep learning theory: generalization, kernel limits, gaussian processes probabilistic models
  • Adversarial robustness
  • Machine learning on graphs
  • Statistical learning theory

Recent News

April 29, 2026Robustness certificates for neural networks against data poisoning and evasion attacks got accepted at IEEE Open Journal of Control Systems
April 15, 2026Out paper Exact Certification of Neural Networks and Partition Aggregation Ensembles against Label Poisoning got awarded a Spotlight (top 6.4%) at Trustworthy AI Workshop, ICLR 2026 🎉
Feb 2, 2026I am joining AIML, Adelaide University as a Research Fellow! Looking forward to the new chapter!
Nov 26, 2025Successfully defended my PhD thesis! 🎓
Aug 5 - Nov 10, 2025Excited to visit Dr. Russell Tsuchida at Monash University for a research visit.
Show more news
June 16, 2025Submitted my PhD dissertation!
May 15, 2025Excited to give a contributed talk on Generalization Certificates for Adversarially Robust Bayesian Linear Regression at the Advances in post-Bayesian methods workshop, UCL, London, U.K.!
Feb 19 – 21, 2025 Excited to give a talk on Certifying (Graph) Neural Networks against Data Poisoning using the Neural Tangent Kernel at understanding generalization in deep learning workshop at Burghausen, Germany!
Feb 11, 2025Our paper Exact Certification of (Graph) Neural Networks Against Label Poisoning got accepted as a Spotlight (top 5.1%) at ICLR 2025 🎉
Dec 5, 2024Our paper Provable Robustness of (Graph) Neural Networks Against Data Poisoning and Backdoor Attacks won the best paper award at AdvML-Frontiers@NeurIPS 2024 🎉
Nov 4, 2024Won 3rd prize in EMCR talk at AI in Science Conference 2024, Canberra, Australia 🏆