Shared Stochastic Gaussian Process Latent Variable Models: A Multi-modal Generative Model for Quasar Spectra

Vidhi Lalchand [email protected], Anna-Christina Eilers [email protected] · 2025

Read Shared Stochastic Gaussian Process Latent Variable Models: A Multi-modal Generative Model for Quasar Spectra by Vidhi Lalchand [email protected], Anna-Christina Eilers [email protected] (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-19824.txt.

Description

Vidhi Lalchand and Anna-Christina Eilers build a shared model for the light spectra of quasars and their known traits. Quasars are very bright galaxies with huge black holes at their cores. Their 2025 paper joins two kinds of data for each object in one model.

The model hides each quasar in a shared low-size space, then rebuilds both its spectrum and its labels from that space. Labels include black hole mass, feeding rate, and brightness. The method can train even when many values for an object are missing. Tests rebuild test spectra and labels with high fidelity. The authors also note what the hidden space may mean for quasar science.

Catalog record

CREATORVidhi Lalchand [email protected], Anna-Christina Eilers [email protected]
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-19824

Browse the full library · About the Tower of Babel