Daattavya Aggarwal

I am currently in the final year of a PhD in AI for Mathematical Discovery at the Computer Lab in Cambridge.

My primary research interests can be broadly summarized through the following question: What is the nature of mathematical discovery and what role can AI play in this process?

I am tackling this question through multiple approaches:

• Guided mathematical discovery: Developing systems that are capable of proposing definitions and asking mathematical questions outside of their training data, that they judge to be `interesting’.
• Impact of proof and formalization: Sparse, verifiable reward has been crucial in developing models that are successful at answering fixed mathematical questions. I am interested in the relationship between informal and formal proof and their roles on posing mathematical questions.
• Interpretability of mathematical thought: Over the past 6 months, frontier reasoning models have demonstrated the capability to resolve open research problems. How do they `think’ while working on mathematical tasks?

I am interested in applying tools from artificial intelligence to a variety of mathematical areas and in the past, I have worked on problems in topology, geometry and number theory. Previously, I completed a masters in mathematical and theoretical physics at Oxford with a focus on Calabi-Yau manifolds.

For an up-to-date list of publications, please visit my Google Scholar profile.

Current situation in AI for Mathematics: I strongly believe in the importance of openness and transparency for the scientific endeavour. For a host of reasons, I would like to work towards a future where there is equitable access to open systems that the mathematical community can choose to engage with.

This website uses the al-folio theme.