Szepesvari walks back Lean doubts: printed axioms plus sorry-free proofs suffice
CsabaSzepesvari · x · 2026-09-07
In the third post of his thread on Anthropic-related Lean formalization, Csaba Szepesvari refines his earlier objection: formalizing a proof introduces many auxiliary definitions and proofs, and one mistakenly added axiom means you're proving something else — but on reflection he thinks his worry was overblown.
Key arguments: all axioms used in a proof can be printed for inspection; a sorry-free proof is a proof apart from Lean bugs; hand-checking even a FLT-scale formalization is feasible. A candid public walkthrough of reasoning about trust in AI-assisted formal mathematics.
Related event: Claude's FLT Proof Ignites Math Community Scooping-Ethics Debate(26 posts)→
More from Research
- FrontierML benchmark shows open models still far behind closed ones on end-to-end ML work — scaling01 · 2026-10-27
- KANLab: an interactive browser playground for Kolmogorov-Arnold Networks — Lukas Niehaus · 2026-10-27
- CMU's LLM2Cert synthesizes stability certificates verified via SMT and Lean — CarnegieMellonU · 2026-10-27
- NVIDIA open-sources FLORA, a robotics data engine ~125x cheaper than GPT-6 Astra — nvidia · 2026-10-27
- RSNA 2026 Deep Learning Labs return: eight hands-on AI classes over four days in Chicago — PRouzrokh · 2026-10-27
- Steering two orthogonal concept neurons makes a model compose "gnome" — Sauers_ · 2026-10-27