AI Co-Mathematician: Accelerating Mathematicians with Agentic AI

2026-05-07

AI Summary

This work proposes the first intelligent mathematical workspace designed to emulate human collaborative patterns, specifically addressing the highly iterative and uncertain nature of mathematical research. Built upon a state-aware AI agent architecture, the platform integrates intention refinement, failure hypothesis tracking, and native mathematical expression generation to support end-to-end collaborative exploration—from idea conception and literature retrieval to theorem proving and theory construction. The system effectively manages uncertainty, recovers overlooked prior work, and aids in identifying novel research directions. Evaluated on the FrontierMath Tier 4 benchmark, it achieves a score of 48%, establishing a new state-of-the-art performance for AI systems and significantly advancing the role of artificial intelligence in open-ended mathematical discovery.

Abstract

We introduce the AI co-mathematician, a workbench for mathematicians to interactively leverage AI agents to pursue open-ended research. The AI co-mathematician is optimized to provide holistic support for the exploratory and iterative reality of mathematical workflows, including ideation, literature search, computational exploration, theorem proving and theory building. By providing an asynchronous, stateful workspace that manages uncertainty, refines user intent, tracks failed hypotheses, and outputs native mathematical artifacts, the system mirrors human collaborative workflows. In early tests, the AI co-mathematician helped researchers solve open problems, identify new research directions, and uncover overlooked literature references. Besides demonstrating a highly interactive paradigm for AI-assisted mathematical discovery, the AI co-mathematician also achieves state of the art results on hard problem-solving benchmarks, including scoring 48% on FrontierMath Tier 4, a new high score among all AI systems evaluated.

Problem

• AI co-mathematician
• mathematical research
• exploratory workflow
• interactive AI
• theorem proving

Innovation

• Agentic AI
• Mathematical reasoning
• Interactive AI workbench
• Stateful collaboration
• Automated theorem proving