Learning JWST. I. A Foundation Model for New Population Discoveries and Morphology-Aware Photometric Redshift Measurements in the JADES Survey
Abstract
We present FM-JADES-v1, a self-supervised foundation model for James Webb Space Telescope ({\em JWST}) deep-field science, trained with 482,444 objects from the {\em JWST} Advanced Deep Extragalactic Survey (JADES) Data Release 5 using multi-band imaging and the photometric catalog. The shared embedding space is trained without class labels. We demonstrate that FM-JADES-v1 can serve as a powerful tool for object discovery and improving property measurements using two experiments, blind active discovery and few-band photometric redshift. For blind object discovery, FM-JADES-v1 identifies rare object populations such as high-redshift galaxies and Little Red Dots (LRDs) without any prior population labels or population-specific selection criteria. These rare populations emerge as isolated islands in the embedding space, which can be identified without prior astrophysical knowledge. For few-band photometric redshift, FM-JADES-v1's learned embeddings achieve σNMAD=0.157 in a strictly controlled three-band (F115W/F200W/F356W) photo-z benchmark, compared to σNMAD= 0.44 for template fitting. These results demonstrate the potential of self-supervised multi-modal representations as scalable discovery spaces for large astronomical surveys. Applied to ongoing and future wide-field surveys from JWST, Roman, Euclid, and Rubin/LSST, this framework could enable systematic searches for rare populations, as well as enabling multiple downstream tasks such as improving astrophysical property measurements.
Related
- Tabular Foundation Models Dramatically Outperform Bespoke Photometric Redshift Estimators
- First tomographic measurements of the angular clustering and bias of photometric quasars from S-PLUS
- Follow-up of SN 2025wny IV: Photometric Time-delay Measurements of a Strongly Lensed Superluminous Supernova
- FOMO-Z. Optimal galaxy classification, photometric redshifts, and reconstruction of population distributions
- Data-driven Galaxy Population Prior for Photometric Redshifts
BibTeX
@misc{ding2026learning,
title = {{Learning JWST. I. A Foundation Model for New Population Discoveries and Morphology-Aware Photometric Redshift Measurements in the JADES Survey}},
author = {Jiani Ding and Minghao Yue and Yongda Zhu and Xiaohui Fan and Yufeng Luo},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.11879},
eprint = {2609.11879},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.11879},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Learning JWST. I. A Foundation Model for New Population Discoveries and Morphology-Aware Photometric Redshift Measurements in the JADES Survey
AU - Jiani Ding
AU - Minghao Yue
AU - Yongda Zhu
AU - Xiaohui Fan
AU - Yufeng Luo
PY - 2026
DA - 2026/09/10
PB - arXiv
DO - 10.48550/arxiv.2609.11879
UR - https://arxiv.org/abs/2609.11879
AB - We present FM-JADES-v1, a self-supervised foundation model for James Webb Space Telescope ({\em JWST}) deep-field science, trained with 482,444 objects from the {\em JWST} Advanced Deep Extragalactic Survey (JADES) Data Release 5 using multi-band imaging and the photometric catalog. The shared embedding space is trained without class labels. We demonstrate that FM-JADES-v1 can serve as a powerful tool for object discovery and improving property measurements using two experiments, blind active discovery and few-band photometric redshift. For blind object discovery, FM-JADES-v1 identifies rare object populations such as high-redshift galaxies and Little Red Dots (LRDs) without any prior population labels or population-specific selection criteria. These rare populations emerge as isolated islands in the embedding space, which can be identified without prior astrophysical knowledge. For few-band photometric redshift, FM-JADES-v1's learned embeddings achieve σ_NMAD=0.157 in a strictly controlled three-band (F115W/F200W/F356W) photo-z benchmark, compared to σ_NMAD= 0.44 for template fitting. These results demonstrate the potential of self-supervised multi-modal representations as scalable discovery spaces for large astronomical surveys. Applied to ongoing and future wide-field surveys from JWST, Roman, Euclid, and Rubin/LSST, this framework could enable systematic searches for rare populations, as well as enabling multiple downstream tasks such as improving astrophysical property measurements.
ER -