LOS mixing & tomography
Different spectra and projected orientations can produce frequency-dependent polarization. Starlight information was already proposed as a remedy.
Tassis & Pavlidou (2015) ↗“Unique among these LLMs” does not mean “new to astrophysics.” The repo contains a useful synthesis and candidate methods, but no sky-data result or validated improvement in an r likelihood.
Different spectra and projected orientations can produce frequency-dependent polarization. Starlight information was already proposed as a remedy.
Tassis & Pavlidou (2015) ↗Moment compression, chromatic E/B and EB, and Planck-compatible 3D complexity already have direct prior art.
Vacher et al. (2023) ↗Vacher et al. (2025) ↗The new notes derive a coherence bound, exact inversion conditioning and a spectral-nullspace test. These clarify the research; priority is unestablished.
Read the derivations →Test whether correctly associated 3D information improves held-out polarization predictions beyond equally flexible empirical and deliberately shuffled priors.
Read the research protocol →A successful result could establish a practical gain from tomography. A null result would constrain which ancillary information is actually useful. Neither outcome is already in this repository.