Researchers Question Lean Verification of GPT's Navier-Stokes Proof

Mathematician Elliot Glazer argues the Lean formalization of GPT's Navier-Stokes proof may not correspond one-to-one with the paper, suggesting shortcuts on hard lemmas. An arXiv paper further contends that translating ambiguous natural-language proofs into Lean is harder than the halting problem, undermining the claimed verification.

2026-10-08 ~ 2026-10-08 · 2 related posts

• ArXiv paper: Lean verification of AI autoformalisation doesn't guarantee correct natural language proofs — RexDouglass · 2026-10-08
• Lean proof may not map 1-to-1 to paper: GPT formalization takes shortcuts on hard lemmas — ctjlewis · 2026-10-08