What was already known
Gaitsgory and Raskin had an equivalence onto a possibly proper open-and-closed union of components, with full support for general linear groups.
AI involvement
OpenAI authors the manuscript, and its release says an internal model produced ordinary evaluation manuscripts. Human task selection and curation are not detailed for this result.
OpenAI authorship and the repository's model-production account support AI-led work; the record does not establish independence from human problem selection.
Importance
Under the stated hypotheses, upgrades the restricted equivalence to the whole spectral category.
Limitations
The theorem relies on the cited earlier equivalence and specialization results. A completed human or machine check of the new full-support argument is undocumented. The later tame eigensheaf theorem is separate.
Technical verification
The released manuscript supplies an argument, but no completed human or machine check of its full-support theorem is documented. Reported is the evidence default, not a failed-check finding.
Human understanding
No named human explains the new support and nonvanishing mechanism; Few is the missing-evidence default.
Correction and publication history
Under the stated hypotheses, upgrades the restricted equivalence to the whole spectral category.
Read this published revision