Axiom

Axiom is the AI For Math company founded by Hong Letong / 洪乐潼 and discussed 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. The source presents the company as a deep-tech systems effort focused on formal mathematical reasoning rather than a generic model company.

Source Position

Axiom’s core product direction is Axiom Prover, an AI theorem-proving system built around Lean Theorem Prover-style verification.
The company is described as needing AI, reinforcement learning, agents, code generation, compiler, Lean, Mathlib, metaprogramming, and pure-math talent.
Hong says Axiom had to build auxiliary tools because Lean data is scarce, the language is brittle, and verification can be slow.
The episode claims Axiom raised a large seed round and then an A round, while emphasizing that investor interest depended on team execution and system progress rather than a simple model demo.
Formal Verification is framed as Axiom’s best first market because chips, software, and complex systems already need stronger proof than finite tests can provide.

Hong Letong / 洪乐潼, Shubo, and Ken Ono — founder, cofounder, and senior mathematician in the source’s Axiom story.
Axiom Prover, AI For Math, AI Mathematician, and Mathematical Abundance — company mission and long-run vision.
Lean Theorem Prover, Mathlib, Interactive Theorem Proving, and Auto-Formalization — formal-methods stack.
Formal Verification, Formal Specification, AI Verification, and AI Coding Verification — commercial and technical verification layer.
AI Organization Design, Subagent Workflow, and AI Skills — operating and system-design patterns the episode connects to Axiom.