Home Zoom Levent Alpöge: From Number Theory to the Riemann Hypothesis and Navier–Stokes
Zoom

Levent Alpöge: From Number Theory to the Riemann Hypothesis and Navier–Stokes

Levent Alpöge: Riemann, Navier–Stokes, Google Scholar

Levent Alpöge, born in 1992, is an American and Turkish mathematician whose career has moved from classical number theory at Harvard and Princeton to the technical staff of Anthropic, the artificial-intelligence company. For most of his career, his name appeared in Inventiones Mathematicae, Combinatorica, and other venues associated with arithmetic statistics and rational points on curves. By late 2026, however, Alpöge had become one of the most closely watched figures in mathematics, not because of a conventional journal article, but because of a series of AI-assisted claims touching the Riemann Hypothesis and the Navier–Stokes existence and smoothness problem.

Those claims must be read with care. Alpöge’s 2026 work is dramatic, but much of it remains at the preprint or formalization stage rather than in the settled literature. The Riemann announcement, in particular, is a Lean-only artifact of extraordinary size; independent kernels accept the archive, but mathematicians have not yet digested a human-readable proof. The Navier–Stokes work, meanwhile, is a technically human-readable preprint with a Lean formalization and an unusually rushed write-up. The contrast between Alpöge’s conventional scholarly record and his recent announcements is itself the story.

You Might Like: The “Dirty Bomb” Meaning

Riemann, Navier–Stokes, and the Verification Question

On 20 November 2026, Levent Alpöge announced that he believed he had found a proof of the Riemann Hypothesis, and that the method extended to the Generalized Riemann Hypothesis for Dirichlet L-functions. The work was produced with Anthropic’s Claude model and presented as a Lean 4 formalization rather than a conventional paper.

Terence Tao described the artifact as “enormous” and “Lean-only,” noting that the proof appears to draw on modern developments in the Langlands program and arithmetic geometry, but that the digestion process “will take some time.” A reader’s guide published the following day reports that Lean’s kernel and two independent kernels accept the archive, yet the absence of a human-readable proof has left the mathematical community unable to assess the argument in the usual way.

The Navier–Stokes episode is more developed but equally contested. With Tristan Buckmaster of New York University, Alpöge produced an AI-assisted proof of finite-time blowup for the unforced incompressible Navier–Stokes equations in three dimensions. The result was framed as closing a loophole in OpenAI’s September 2026 solution, which relied on an artificial forcing term that many experts considered unrealistic.

Buckmaster and Alpöge’s preprint has a Lean formalization, but the write-up was released in an unpolished state and initially lacked the certificate of the result; the authors said they could not risk waiting because of the earlier scooping controversy. That controversy, in which Buckmaster alleged that OpenAI had accessed his private work, remains part of the background against which the new proof must be judged.

The Google Scholar Trail of a Traditional Mathematician

A Google Scholar search for Levent Alpöge still returns primarily his pre-2026 number-theory and combinatorics papers: “Self-conjugate core partitions and modular forms” in the Journal of Number Theory, “Proof of a conjecture of Stanley–Zanello” in the Journal of Combinatorial Theory, Series A, and, more recently, “Rank stability in quadratic extensions and Hilbert’s tenth problem for the ring of integers of a number field” in Inventiones Mathematicae. These are conventional, peer-reviewed works with citation counts in the low hundreds across public databases—a respectable record for a mathematician trained under Manjul Bhargava, but a record that says nothing about the AI-assisted claims of 2026.

Also See: How Old Was Daniel Craig in Casino Royale?

That gap is the crucial point. Google Scholar indexes journal articles and stable preprints; it does not yet reflect the Lean archives, formalization repositories, and rapid announcements that now define Alpöge’s public profile. His scholarly footprint remains that of a careful arithmetic geometer, while his current reputation rests on results whose verification is still unfolding. Whether the Riemann and Navier–Stokes claims eventually enter the indexed literature—or remain striking but unassimilated formal artifacts—will determine how the next chapter of that Google Scholar record is written.