Researchers Resolved Six Erdős Mathematical Problems
The awardees utilized AI-assisted formal verification to crack challenges from a catalog of over 1,200 entries.
Updated on Oct. 5, 2026 in Mathematics
Live Poll
Do you trust AI-assisted tools to improve the accuracy of scientific discoveries?
The Office of Justin Sun has awarded the Justin Sun Prize to Wouter van Doorn, Quanyu Tang, and Yanyang Li for solving six problems from the Erdős catalog. These findings leveraged a combination of human judgment and AI-based proof tools.
Why it matters
The awards aim to accelerate the intersection of mathematical research, formal verification, and AI-assisted scientific discovery. By utilizing computer-aided tools, the researchers have begun to address a massive backlog of over 1,200 open challenges.
The researchers resolved six problems from the Erdős catalog, which contains more than 1,200 items. Van Doorn verified three problems in Lean—a proof assistant software that ensures mathematical correctness—while other team members utilized the AI system Aristotle.
The players
Wouter van Doorn
A mathematician whose research spans the application of Lean formal verification tools to complex problems.
The Office of Justin Sun
An organization focused on promoting advancements in mathematics and decentralized technology that administers the prize.
The details
The researchers employed a hybrid approach combining human expertise with computational proof strategies. Wouter van Doorn produced computer-checkable proofs in Lean, a software environment designed for formal verification, for problems #369, #457, and #469. For problem #650, the team used ChatGPT to develop a strategy, while the AI system Aristotle, a custom tool for automated reasoning, was used to repair a gap in formalization. Tang resolved problem #1044, while Tang and Li contributed to the resolution of problem #1196.
Timeline
2010: Wouter van Doorn began his undergraduate mathematics research.
October 5, 2026: The Office of Justin Sun announced the prize recipients in Geneva, Switzerland.
The Tech Race
The effort to systematically clear the Erdős catalog represents a significant push to automate mathematical verification at scale. This project follows a trend of using AI proof assistants to resolve long-standing problems that previously resisted manual efforts.
The project demonstrates the practical utility of formal verification tools for researchers working on complex logical proofs. While not a consumer-facing application, the techniques deployed here provide a framework for future AI-assisted scientific discovery across multiple domains.
The takeaway
This success shows that hybrid human-AI workflows are capable of clearing complex backlogs in formal mathematics. Researchers and practitioners should watch for future updates to the Erdős catalog as automated proof methods continue to gain traction.
Further reading
Explore the latest developments in algorithmic research on our Mathematics page.
More information
View complete details regarding the awards and methodology at the Justin Sun Prize information page.
Live Poll
Do you trust AI-assisted tools to improve the accuracy of scientific discoveries?
