I’m a computer science and biology student at the University of British Columbia, learning interpretability in Xin Tang’s labXin Tang’s labAt UBC. Our group meets weekly to take apart papers on how AI scientists form and test ideas.lab site, where our group looks inside AI scientists to see how they judge ideas.
I build at Antimatters Labs, a small non-profit working on AI for science: models that keep the known physics and learn the rest, and tools for research agents like scimapscimapInfrastructure for autonomous research: one record every agent run writes into. swafra builds on it, with 94.7% recall on LongMemEval.sitecodeswafra, iceiceVideo-call an AI scientist while it runs experiments, with 2,000+ biology tools.democode and Dry-LabDry-LabAgents take raw biological data to a cited report, a person approving each step.democode. Docking drugs for Parkinson’s there drew me to biology and the Longevity Biotech Fellowship. Before that, I modelled microgrids with neural ODEs alongside Dr. Raj DandekarDr. Raj DandekarMIT PhD and co-founder of Vizuara; works on scientific machine learning.scholarour paper.
Before moving to Canada, I taught physics to my classmates, and later IELTS and PTE. At 17 I started Antimatters Technology, tried chip design in a small lab I set up, and built software for immigration startups in my hometown. In Canada, I finished Seneca College’s three-year programming diploma in two years, with two internships.