- DPhil student @ Intelligent Earth CDT in AI for the Environment 🌎
- GOALS Group @ Oxford Robotics Institute 🤖
- Safe‑exploration RL for disaster response ⛑️
- Building open‑source tools & dashboards that connect ML theory with impact
-
RUBIX Rubix is a fully tested, well-documented, and modular Open Source tool developed in JAX, designed to forward model IFU cubes of galaxies from cosmological hydrodynamical simulations. The code automatically parallelizes computations across multiple GPUs, demonstrating performance improvements over state-of-the-art codes by a factor of 600. This optimization reduces compute times from hours to only seconds. RUBIX leverages JAX’s auto-differentiation capabilities to enable not only forward modeling but also gradient computations through the entire pipeline paving the way for new methodological approaches such as e.g. gradient-based optimization of astrophysics model parameters.
-
MEGS
Morphological Evaluation of Galactic Structure (MEGS) implements PCA in Python to compress mass, metallicity and stellar‑age maps from the IllustrisTNG simulation into a low‑dimensional feature space that reflects true morphological similarity. It reduces 2D image dimensions by ~200x and 3D cube dimensions by ~3650x while holding reconstruction error below 5%. MEGS delivers an interpretable generative model for galaxies, ready for downstream tasks such as classification and structural analysis. -
ESO: Evolutionary Spectrogram Optimization for Passive Accoustic Monitoring ESO applies a genetic algorithm to identify the most informative spectrogram regions for passive acoustic monitoring. It cuts model complexity by 91% and inference time by 70% while retaining 96% of the original F1 score.
-
GAMMA Dataset Public dataset of mass, metallicity & age attributes for galaxies (Zenodo).