Closing the stellar labels gap: Stellar label independent evidence for [$\alpha/M$] information in Gaia BP/RP spectra

TL;DR

This paper introduces a novel label-independent model using a variational auto-encoder to analyze Gaia BP/RP spectra, effectively capturing stellar properties like [$ extalpha$/M] without relying on stellar labels, thus addressing the stellar labels gap.

Contribution

The authors develop a label-independent variational auto-encoder model that accurately reconstructs Gaia XP spectra and detects [$ extalpha$/M] information without label dependence, advancing stellar spectral analysis.

Findings

Model achieves competitive spectral reconstructions.

Successfully detects [$ extalpha$/M] bimodality.

Sensitive to anomalous Gaia-Enceladus star abundances.

Abstract

Data-driven models for stellar spectra which depend on stellar labels suffer from label systematics which decrease model performance: the "stellar labels gap". To close the stellar labels gap, we present a stellar label independent model for Gaia BP/RP spectra. We develop a novel implementation of a variational auto-encoder, which learns to generate an XP spectrum and accompanying 'scatter' without relying on stellar labels. We demonstrate that our model achieves competitive XP spectra reconstructions in comparison to stellar label dependent models. We find that our model learns stellar properties directly from the data itself. We then apply our model to XP/APOGEE giant stars to study the [$\alpha$/M] information in Gaia XP. We provide strong evidence that the XP spectra contain meaningful [$\alpha$/M] information by demonstrating that our model learns the $\alpha$-bimodality, without relying on stellar label correlations for stars with $T_{\rm eff} <$ 5000 K, while also being sensitive to the anomalous abundances of Gaia-Enceladus stars. We publicly release our trained model, codebase and data. Importantly, our stellar label independent model can be implemented for any/all XP spectra because our model performance scales with training object density, not training label density.