English

Spatio-Spectroscopic Representation Learning using Unsupervised Convolutional Long-Short Term Memory Networks

Astrophysics of Galaxies 2026-02-23 v1 Computer Vision and Pattern Recognition

Abstract

Integral Field Spectroscopy (IFS) surveys offer a unique new landscape in which to learn in both spatial and spectroscopic dimensions and could help uncover previously unknown insights into galaxy evolution. In this work, we demonstrate a new unsupervised deep learning framework using Convolutional Long-Short Term Memory Network Autoencoders to encode generalized feature representations across both spatial and spectroscopic dimensions spanning 1919 optical emission lines (3800A <λ<< \lambda < 8000A) among a sample of 9000\sim 9000 galaxies from the MaNGA IFS survey. As a demonstrative exercise, we assess our model on a sample of 290290 Active Galactic Nuclei (AGN) and highlight scientifically interesting characteristics of some highly anomalous AGN.

Keywords

Cite

@article{arxiv.2602.18426,
  title  = {Spatio-Spectroscopic Representation Learning using Unsupervised Convolutional Long-Short Term Memory Networks},
  author = {Kameswara Bharadwaj Mantha and Lucy Fortson and Ramanakumar Sankar and Claudia Scarlata and Chris Lintott and Sandor Kruk and Mike Walmsley and Hugh Dickinson and Karen Masters and Brooke Simmons and Rebecca Smethurst},
  journal= {arXiv preprint arXiv:2602.18426},
  year   = {2026}
}

Comments

This manuscript was previously submitted to ICML for peer review. Reviewers noted that while the underlying VAE-based architecture builds on established methods, its application to spatially-resolved IFS data is promising for unsupervised representation learning in astronomy. This version is released for community visibility. Reviewer decisions: Weak accept and Weak reject (Final: Reject)