I am a second-year PhD student at the School of Engineering in the University of Edinburgh, supervised by Dr. Elliot J. Crowley and Dr. Shay Cohen. I am a member of the BayesWatch group.

My research focuses on the efficient training of Large Language Models (LLM) as well as their applications in Neural Architecture Search (NAS).

Previously, I received my First Class MSci degree in Computer Science (Artificial Intelligence) from King’s College London.

News

[10/2025] Released our work ONNX-Net, a framework for universal representations and instant performance prediction in neural architecture search. Checkout the preprint, project page and our dataset.
[05/2025] Our paper on Transferrable Surrogates accepted to AutoML 2025!
[04/2025] Released our work on Transferrable Surrogates for NAS. Checkout the preprint and our project page.

Publications

ONNX-Net: Towards Universal Representations and Instant Performance Prediction for Neural Architectures[preprint][project page][dataset][code] Shiwen Qin, Alexander Auras, Shay B. Cohen, Elliot J. Crowley, Michael Moeller, Linus Ericsson, Jovita Lukasik.

Transferrable Surrogates in Expressive Neural Architecture Search Spaces [paper][project page][code] Shiwen Qin, Gabriela Kadlecová, Martin Pilát, Shay B. Cohen, Roman Neruda, Elliot J. Crowley, Jovita Lukasik, Linus Ericsson, in AutoML, 2025.

Teaching

I have taken the role of demonstrating and marking while studying in Edinburgh. Courses include Data Analysis and Machine Learning and Introduction to Python.