Provenance Record

Tightly written 710-word News piece with a clear Overview / What We Know / What We Don't Know / Analysis structure. Technical depth is appropriate and accessible; the inversion finding is explained plainly without overreach.

Both source snapshots read from disk (gunzip). source-0.html.gz (nature.com, status 200) — verified the title 'Cortex-specific inversion of visual responses during sleep', DOI 10.1038/s41467-026-73997-y verbatim, publication date 11 June 2026, and all five authors (Nicholas G. Cicero, Michaela Klimova, Louis Vinke, Sam Ling, Laura D. Lewis) plus affiliations (Boston University, MIT, Mass General Psychiatry, Martinos Center). All five direct quotes appear verbatim in the abstract: 'responses to light remained intact in the visual thalamus during N1 and N2 sleep'; 'stimulus-evoked responses in early visual cortex were profoundly suppressed'; 'sensory information can still reach the cortex'; 'an inverted pattern in which high-intensity visual stimulation evoked visual cortical deactivation'; 'These findings suggest a cortical mechanism where inhibitory circuits regulate stimulus-driven deactivation in visual cortex, facilitating sensory isolation during early stages of sleep'. source-1.html.gz (sites.bu.edu, status 200) — the Sam Ling Vision Lab publications page lists the paper among its 2026 publications with the exact author string and 'Nature Communications', corroborating the secondary claim only.

Every technical specific traces to the Nature Communications paper (Rule 9): headline, lead, DOI, date, authors, the thalamus-intact / cortex-inverts split, and the inhibitory-gating interpretation (correctly framed as the authors' proposal, not established fact). No participant count and no BOLD/percent-signal/p-value statistics were stated — confirmed none were invented (Rule 2). The DOI string is correct and not fabricated. Direction of the effect (cortical deactivation to bright stimuli) is reported accurately.

High-quality, fully sourced submission. The central thesis is entirely carried by the primary Nature Communications paper; the BU lab page is a legitimate corroborating institutional source whose registrable domain (bu.edu) is already allowlisted. The single warning is a subdomain-matching artifact, not a factual or attribution defect, so there is nothing for a corrections record to honestly document. APPROVE.