Scalar filters preserved spectral loop phases in passive models of two partial adult worm maps

We tested a spectral loop check in reciprocal passive models of two partial adult worm electrical maps. Each model had 180 cells, with 60 observed cells. Scalar low-pass filters, gains and delays changed pair phases but preserved triangle loop phases to within 7e-16. Equal independent source noise gave a zero loop sine. At a noise variance ratio bounded by two and angular frequency one, the median mean absolute loop sine over 50 noise patterns was 0.01824 and 0.01366. A deterministic 1% relative cross-spectrum error bound certified a non-real loop product in median shares of 18.87% and 10.77% of qualifying triangles. The error bound was added after the first run. A separate three-cell example had irreversible dynamics but zero loop phase. Thus nonzero loop phase can reject a reversible latent process under diagonal filtering, but zero loop phase cannot establish reversibility. These are exact population model results, not measured worm activity or finite-record detection rates. Partial anatomy, assumed conductances and noise, one observed set, and absent activity data limit biological use. Original code, aggregate results, a figure and external retrieval instructions support reproduction.

connectomics · spectral phase · closure phase · time reversibility · calcium filters · C. elegans