Projects alongside my main research.
Sep 2022 – Nov 2022
Implemented quantum Hadamard edge detection (QHED; Yao et al., 2017): images amplitude-encoded on qubits, with a Hadamard gate on an auxiliary qubit computing differences between neighbouring pixels to extract horizontal and vertical edges.
Aug 2023 – Oct 2023
Metropolis sampling with simulated annealing to form topologically protected skyrmion configurations on a 2D square lattice.
May 2022 – Jul 2022
Simulated a CsI(Tl) scintillator detector setup in Geant4, the C++ Monte Carlo toolkit for the passage of particles through matter. Wrote the detector construction in C++ and counted the hits generated on the detector surface, the input to the detector’s pulse-height spectrum.
Feb 2024 – Mar 2024
Implemented the ellipsoid and EXP4 contextual-bandit algorithms of Cohen et al. (2020) and compared their regret, and how they adapt, under changing distributions.
Aug 2023 – Oct 2023
A Streamlit app that answers questions over multiple uploaded PDFs, using agentic chunking into propositions with Google Gemini and Hypothetical Document Embeddings (HyDE) to improve retrieval.
Mar 2023 – Apr 2023
Trained a conditional GAN for image-to-image translation in the Lab colour space on 8,000+ COCO images.
Jun 2022 – Jul 2022
Implemented a deep Q-network (DQN) in PyTorch that learns to act directly from raw game frames through a convolutional network; the trained agent clears the first level of Super Mario Bros.