University researcher awarded $1.5m to advance AI in mathematics

University of Melbourne Professor Cezary Kaliszyk is among the first cohort to be awarded a $1.5 million (USD$1m) grant by the international AI for Math Fund, which seeks to advance the impact of math discovery globally.

Based in the School of Computing and Information Systems, Professor Kaliszyk will collaborate with Drs Martin Suda and Jan Jakubův at the Czech Technical University in Prague. The team’s DEEPER project centres on developing AI-based guidance for automated reasoning systems, designing efficient proof automation, and integrating these advancements into modern ‘proof assistants’ for mathematicians and other experts.

They will seek to do this by combining two distinct approaches to artificial intelligence: neural learning, which relies on pattern recognition in data; and symbolic learning, which interprets explicit data. In doing so, this project aims to bridge gaps in existing models for a more well-rounded proof assistant that can both interpret and extrapolate data autonomously.

“We are delighted to receive this grant,” Professor Kaliszyk said.

“Advances in AI reasoning and neural learning have huge potential for accelerating mathematical research and science, and could benefit sectors from health to cybersecurity to manufacturing, and much more.”

Another crucial component of the project is to develop a more robust verification process for mathematical proofs. Through integration and enhancement of existing proving systems, the team’s work, grounded in formal logic, seeks to ensure every result meets the highest standard of mathematical rigor.

The AI for Math Fund is a joint initiative of Renaissance Philanthropy and XTX Markets, with the DEEPER project selected from over 280 applications across the globe.