Autoencoding Galaxy Spectra I: Architecture

Abstract

We introduce the neural network architecture SPENDER as a core differentiable building block for analyzing, representing, and creating galaxy spectra. It combines a convolutional encoder, which pays attention to up to 256 spectral features and compresses them into a low-dimensional latent space, with a decoder that generates a restframe representation, whose spectral range and resolution exceeds that of the observing instrument. The decoder is followed by explicit redshift, resampling, and convolution transformations to match the observations. The architecture takes galaxy spectra at arbitrary redshifts and is robust to glitches like residuals of the skyline subtraction, so that spectra from a large survey can be ingested directly without additional preprocessing. We demonstrate the performance of SPENDER by training on the entire spectroscopic galaxy sample of SDSS-II; show its ability to create highly accurate reconstructions with substantially reduced noise; perform deconvolution and oversampling for a super-resolution model that resolves the [OII] doublet; introduce a novel method to interpret attention weights as proxies for important spectral features; and infer the main degrees of freedom represented in the latent space. We conclude with a discussion of future improvements and applications.

17 pages, 5 figures, submitted to AJ. Code available at https://github.com/pmelchior/spender

References in corpus (7)

• Array Programming with NumPy
• EAZY: A Fast, Public Photometric Redshift Code
• Improving the full spectrum fitting method: accurate convolution with Gauss-Hermite functions
• Understanding Galaxy Evolution through Emission Lines
• The Completed SDSS-IV extended Baryon Oscillation Spectroscopic Survey: Large-scale Structure Catalogs for Cosmological Analysis
• Autoencoding Galaxy Spectra II: Redshift Invariance and Outlier Detection
• Mapping the Diversity of Galaxy Spectra with Deep Unsupervised Machine Learning

Cited by in corpus (14)

• AstroCLIP: A Cross-Modal Foundation Model for Galaxies
• Autoencoding Galaxy Spectra II: Redshift Invariance and Outlier Detection
• The ROAD to discovery: machine learning-driven anomaly detection in radio astronomy spectrograms
• Deriving the star formation histories of galaxies from spectra with simulation-based inference
• Impact of stellar population synthesis choices on forward modelling-based redshift distribution estimates
• Inferring redshift and galaxy properties via a multi-task neural net with probabilistic outputs: An application to simulated MOONS spectra
• Galaxy Spectra Networks (GaSNet). III. Generative pre-trained network for spectrum reconstruction, redshift estimate and anomaly detection
• Ly$\alpha$NNA: A Deep Learning Field-level Inference Machine for the Lyman-$\alpha$ Forest
• SpecPT (Spectroscopy Pre-trained Transformer) Model for Extragalactic Spectroscopy: I. Architecture and Automated Redshift Measurement
• Automated quasar continuum estimation using neural networks: a comparative study of deep-learning architectures
• QUEST (Quasar Unsupervised Encoder and Synthesis Tool): A machine learning framework to generate quasar spectra
• GalSBI: Phenomenological galaxy population model for cosmology using simulation-based inference
• Towards model-free stellar chemical abundances. Potential applications in the search for chemically peculiar stars in large spectroscopic surveys
• Compression Method for Solar Polarization Spectra Collected from Hinode SOT/SP Observations