Return to Colloquia & Seminar listing
Benchmarking to Infinity
Mathematics of Data & Decisions| Speaker: | Dr. Elliot Glazer, Principia Labs. |
| Related Webpage: | https://www.principialabs.org/ |
| Location: | 1025 Physical and Data Science Building |
| Start time: | Tue, Jan 13 2026, 3:10PM |
I will discuss FrontierMath, a mathematical problem solving benchmark I developed last year, including its design philosophy (and flaws) and what we have learned about AI's trajectory from it. I will then look much further out, speculate about what a "perfectly efficient" mathematical intelligence should be capable of, and discuss how high-ceiling math capability metrics can illuminate the path towards that ideal.