Hello there. Welcome to my website.

A little bit about myself :

• I’m a Research Engineer at Google DeepMind, at the Frontier AI team, where I work on continual learning research.
• [NEWS] Our new paper Feedforward Mixing is as Sharp as it is Slow in Reverse has been accepted to the ICML 2026 Workshop on Combining Theory and Benchmarks.
• [NEWS] T-GRAB has been accepted to the KDD 2026 Data & Benchmarks (D&B) track! (originally presented as an oral presentation at the KDD 2025 MLoG-GenAI workshop).
• In the past, I’ve held multiple visiting researcher appointments at University of Oxford, University of Cambridge, MILA, and EMBL-EBI.
• I did my Master’s at the University of Oxford, where I was affiliated with the Oxford Applied Theoretical Machine Learning Group (OATML), supervised by Prof. Yarin Gal. My thesis, “Fine-Tuning Large Language Models to Abstain Appropriately via Semantic Entropy” was recently accepted to NeurIPS Safe Generative AI Workshop 2024. Additionally, I collaborated with Prof. Michael Bronsteinon temporal graph learning. Our work “Enhancing the Expressivity of Temporal Graph Networks via Source-Target Identification” was accepted to NeurIPS Symmetry and Geometry in Neural Representations Workshop 2024 as an oral presentation.
• Prior to Oxford, I did my Bachelor’s at the University of Cambridge, where I was the valedictorian of my cohort.
• Other than machine learning, I have extensive experience (5 years) in engineering, having spent time at high-frequency trading firm XTX Markets and Amazon Web Services.