Contrastive Astromer: Two-Stage Contrastive Representation Learning for Astronomical Light Curve Classification
Abstract
Large astronomical surveys collect abundant light curves, but reliable class labels remain scarce. ASTROMER 2 learns from unlabeled light curves by reconstructing masked observations, yet this objective does not explicitly encourage class separation in a compact representation of each light curve. We introduce Contrastive Astromer, a two-stage approach that makes these representations more useful for classification without changing the ASTROMER 2 encoder architecture. First, unsupervised contrastive learning on single-band MACHO light curves brings together representations of augmented views of the same object. Supervised contrastive learning, supported by an auxiliary classification loss, then uses a separate labeled Alcock sample to strengthen class discrimination. We evaluate the resulting representations by freezing the encoder and training a separate classifier. On a held-out set of 600 objects, this classifier achieves a macro-F1 of 0.7943 with 500 labeled examples per class, exceeding our reproduction of the ASTROMER 2 finetune-and-classify pipeline by 5.10 percentage points. With downstream classifier training budgets of 20 and 100 examples per class, macro-F1 reaches 0.7078 and 0.7704, respectively, also surpassing this baseline. Embedding-space analyses further show that supervised adaptation improves local class consistency and global class separation. These results show that contrastive adaptation makes pretrained light-curve representations more effective for classification and more informative about the structure of variability classes.
Related
- Feature-Based and Deep Learning Methods for Astronomical Light Curve Classification
- Multi-Scale Contrastive Attention for Light-Curve Representation Learning
- Multi-Method Machine Learning Classification of Astronomical Sources with Rubin Data Previews: a Spectroscopy-Free Approach with DP1 and DP2
- Data representation matters: Optimising machine learning for X-ray source classification by leveraging spatio-spectral information
- Data Representation Matters: Optimizing Machine Learning for X-ray Source Classification by Leveraging Spatio-Spectral Information
BibTeX
@misc{wei2026contrastive,
title = {{Contrastive Astromer: Two-Stage Contrastive Representation Learning for Astronomical Light Curve Classification}},
author = {Justin Minggan Wei and Pavlos Protopapas and Guillermo Cabrera-Vives},
howpublished = {arXiv},
year = {2026},
month = nov,
doi = {10.48550/arxiv.2611.00150},
eprint = {2611.00150},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2611.00150},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Contrastive Astromer: Two-Stage Contrastive Representation Learning for Astronomical Light Curve Classification
AU - Justin Minggan Wei
AU - Pavlos Protopapas
AU - Guillermo Cabrera-Vives
PY - 2026
DA - 2026/11/01
PB - arXiv
DO - 10.48550/arxiv.2611.00150
UR - https://arxiv.org/abs/2611.00150
AB - Large astronomical surveys collect abundant light curves, but reliable class labels remain scarce. ASTROMER 2 learns from unlabeled light curves by reconstructing masked observations, yet this objective does not explicitly encourage class separation in a compact representation of each light curve. We introduce Contrastive Astromer, a two-stage approach that makes these representations more useful for classification without changing the ASTROMER 2 encoder architecture. First, unsupervised contrastive learning on single-band MACHO light curves brings together representations of augmented views of the same object. Supervised contrastive learning, supported by an auxiliary classification loss, then uses a separate labeled Alcock sample to strengthen class discrimination. We evaluate the resulting representations by freezing the encoder and training a separate classifier. On a held-out set of 600 objects, this classifier achieves a macro-F1 of 0.7943 with 500 labeled examples per class, exceeding our reproduction of the ASTROMER 2 finetune-and-classify pipeline by 5.10 percentage points. With downstream classifier training budgets of 20 and 100 examples per class, macro-F1 reaches 0.7078 and 0.7704, respectively, also surpassing this baseline. Embedding-space analyses further show that supervised adaptation improves local class consistency and global class separation. These results show that contrastive adaptation makes pretrained light-curve representations more effective for classification and more informative about the structure of variability classes.
ER -