OpenAI reports a regular-lifespan Boltzmann–Grad limit for stable radial potentials

For stable finite-range radial pair potentials in the stated regularity class, grand-canonical Newtonian gases started from spatially summable C¹ data with Gaussian density and spatial-gradient bounds have every fixed scaled factorial marginal converge in full phase-space L¹, uniformly on any finite interval where the corresponding nonlinear Boltzmann equation has a nonnegative classical solution with the required Gaussian bound.

Sources for this record

What was already known

Lanford proved a short-time hard-sphere limit; Pulvirenti–Saffirio–Simonella obtained a short-time theorem for the stable radial class allowing attractive parts. Deng–Hani–Ma reached every regular finite kinetic interval for hard spheres and identified smooth potentials as a further problem (manuscript Section 1.1, p. 4).

AI involvement AI-led

OpenAI credits itself as manuscript author and states that its internal frontier model produced the results in this release. Individual human input is not detailed enough to establish task-only autonomy.

OpenAI is named as author on the manuscript, and its release attributes the results to an internal frontier model. The individual human input is not detailed enough to establish autonomy.

Importance Notable

Extends the regular-lifespan Boltzmann–Grad limit from hard spheres to the stable radial potential class, including attractive wells and noninjective scattering maps.

Limitations

The time interval is conditional on a regular, Gaussian-bounded kinetic solution; the theorem assumes the specified stable finite-range radial potential and grand-canonical excluded-pair initial law. It does not assert global existence of the Boltzmann solution.

Technical verification Reported

Reported is the default when evidence for a higher level is insufficient. The OpenAI-authored manuscript reports and argues the result, but the reviewed sources do not document a completed check of the full assessed claim.

Human understanding Few

Few is the default when evidence for a higher level is insufficient. No named person is documented as explaining the central argument for this assessed contribution.

Correction and publication history

2026-10-18 · publication

Primary manuscript, attribution, and prior-work review for non-notifying local inclusion.

Read this published revision