Universal Spectral Tokenization via Self-Supervised Panchromatic Representation Learning

TL;DR

This paper introduces a self-supervised deep learning model that unifies heterogeneous astronomical spectra into a common, physically meaningful representation, enabling cross-domain analysis and serving as a foundation for scientific models.

Contribution

The authors develop a universal spectral tokenizer that processes diverse spectra directly on native grids, unifying data across resolutions and domains in a self-supervised manner.

Findings

Model produces aligned, homogeneous spectral representations.

Achieves competitive performance on various downstream tasks.

Demonstrates unification of spectral data across resolutions and domains.

Abstract

Sequential scientific data span many resolutions and domains, and unifying them into a common representation is a key step toward developing foundation models for the sciences. Astronomical spectra exemplify this challenge: massive surveys have collected millions of spectra across a wide range of wavelengths and resolutions, yet analyses remain fragmented across spectral domains (e.g., optical vs. infrared) and object types (e.g., stars vs. galaxies), limiting the ability to pool information across datasets. We present a deep learning model that jointly learns from heterogeneous spectra in a self-supervised manner. Our universal spectral tokenizer processes spectra from a variety of object types and resolutions directly on their native wavelength grids, producing intrinsically aligned, homogeneous, and physically meaningful representations that can be efficiently adapted to achieve competitive performance across a range of downstream tasks. For the first time, we demonstrate that a single model can unify spectral data across resolutions and domains, suggesting that our model can serve as a powerful building block for foundation models in astronomy -- and potentially extend to other scientific domains with heterogeneous sequential data, such as climate and healthcare.