Putnam Competition

The Putnam Competition is the undergraduate mathematics competition used in 137. 对洪乐潼的4小时访谈：AI for Math、把数学变成Lean、数学天书中的证明、直觉、被创造与被发现的 as a benchmark for Axiom Prover. The source claims that Axiom formalized the 2025 exam problems and had the system complete all 12 problems for a full score.

Source Position

• The episode treats the Putnam result as a visible milestone, not the whole definition of AI For Math.
• Hong uses it to explain the gap between human elegance and machine proofs: a correct AI proof can be much longer and less intuitive than a human solution.
• The result also functions as an organization story because the team had to formalize problems, run the system, and finish under time pressure.

Connections

• Axiom, Axiom Prover, and Hong Letong / 洪乐潼 — company, system, and founder tied to the source’s Putnam story.
• Lean Theorem Prover, Interactive Theorem Proving, and Auto-Formalization — formalization and proof stack behind the benchmark.
• AI Mathematician and Mathematical Abundance — broader capability claims the source distinguishes from contest success.