GPT-6.2 Astra Formalizes Smoktunowicz's Gap Theorem
Blog post from Vals
On Autoform-Hard’s 56 research-level formalization tasks, GPT-6.2 Astra was the only model to solve a task beyond the 48 completed by GPT-6.2 Sol and Claude Fable 5.2: Agata Smoktunowicz’s theorem that certain finitely generated graded algebras cannot have Gelfand–Kirillov dimension strictly between 2 and 3. Working autonomously in Lean for 94 minutes, Astra produced a 1,794-line proof that passed clean-room grading and independent kernel replay, with no gaps, warnings, or nonstandard axioms. The proof shares some preliminary steps with published arguments but takes a different route for the central result: it defines a “capacity” measuring how much subspaces expand under multiplication, then uses a plateau in that quantity to show that subquadratic growth forces linear growth, contradicting the assumed intermediate dimension. This approach avoids several deep results used in the published proof, including results not available in Lean’s Mathlib, and draws on ideas related to lower transcendence degree and isoperimetric methods.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| LLM | 1 | 3,674 | 823 | 199 | -44% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.