Voiculescu's question: is free entropy dimension an invariant of the generated von Neumann algebra?
Voiculescu's free entropy dimension of a self-adjoint tuple in a tracial von Neumann algebra, together with its variants , and , equals for a free semicircular -tuple generating . If depended only on the von Neumann algebra generated, it would distinguish the free group factors. Does the free entropy dimension of a finite self-adjoint generating tuple depend only on the tracial von Neumann algebra it generates?
- Result
- Disproved(see note)
- Status
- Candidate (review pending)
- AI contribution
- AI-discovered
- Method
- Construction
- Field
- Operator algebras: free probability, free entropy
- Posed by
- Dan Voiculescu (Free entropy, Bull. London Math. Soc. 2002, Section 2.6)
- Year posed
- 2002
- Years open
- 24y
- Solved
- 2026-09-23
- Model
- Unreleased internal OpenAI model
- Vendor
- OpenAI
- Collaborators
- —
- Verification
- Unreviewed
- Publication
- Announced
- Collection
- OpenAI math release (October 2026), version adc7f12
- Significance
- 38 / 100
- Disclosed cost
- —
- Wikipedia
- No dedicated article
What was actually shown
Claims a negative answer for , , and : each takes every integer value on finite self-adjoint generating tuples of the single factor , obtained by transporting semicircular and group-algebra tuples through the new isomorphisms. It does not compute free entropy dimension for other algebras or address the microstates-free equality questions.
What the AI did
The OpenAI math release (github.com/openai/math) states that its results were produced by an unreleased internal OpenAI model, with on average about three hours of ChatGPT Pro thinking compute per result, under one fixed procedure applied to roughly 4,000 posed problems; outputs were then grouped into families and filtered for significance. The manuscript is credited to OpenAI alone and names no human author. This entry is a corollary in the free group factor manuscript.
Verification
No independent mathematician has checked this yet. Checked here: Corollary (dependence on von Neumann generators) of the TeX source read against Voiculescu's question; for every it gives generating tuples of with and others with , a negative answer for all four dimensions. It is a short transport argument that depends entirely on the unrefereed isomorphism theorem. The Lean comparator for the family does not state this corollary.