NAPTIME: A Neural-process Framework for Rubin Alert Classification
4 affiliations
- 1University of Illinois Urbana-Champaign
- 2Center for Astrophysical Surveys
- 3University of Delaware
- 4NSF-Simons AI Institute for the Sky (SkAI)
Abstract
The Vera C. Rubin Observatory Legacy Survey of Space and Time will produce a high-volume stream of irregularly sampled multiband alerts for which spectroscopic confirmation will be available only for a small minority of sources. Tidal disruption events are rare phenomena that provide a direct probe of dormant massive black holes, but their light curves can be confused with nuclear variability and other transient subclasses. We present NAPTIME (Neural Astrophysical Photometric Transient Identification and Modeling Engine), a neural-process framework for photometric transient classification under sparse and partial observational context. NAPTIME models irregular multiband light curves directly, combining probabilistic light-curve reconstruction with classification and optional host-galaxy context, as well as photometric-redshift information. We evaluate on two simulated benchmarks: ELAsTiCC2, our primary Rubin-like broad-classification benchmark, and MALLORN, a photometry-only tidal disruption event (TDE) focused benchmark. On the 15 family ELAsTiCC2 task, the metadata-aware model reaches macro F1 = 0.903 and macro area under the receiver operating characteristic curve (AUROC) = 0.991, while a matched photometry-only variant reaches 0.874 and 0.986. Viewed as a TDE-versus-rest ranking model, the classifier yields TDE average precision 0.985 with metadata and 0.979 without. Metadata is most valuable in the low-context regime. Using only the earliest 10% of retained photometric measurements, macro F1 is ∼0.42 with metadata and ∼0.34 without it. On MALLORN, NAPTIME reaches macro F1 = 0.693 and macro AUROC = 0.958. These results show that neural processes provide a practical probabilistic framework for Rubin-like transient classification and remain effective for TDE-focused candidate recovery.
Versions
Related
- Multi-Method Machine Learning Classification of Astronomical Sources with Rubin Data Previews: a Spectroscopy-Free Approach with DP1 and DP2
- Unresolved Binary Systems in the Rubin Era I: An Autoencoder Framework for Binary Identification Applied to 47 Tucanae
- Modeling Multi-band Active Galactic Nuclei Lightcurves Using an Self-supervised Attentive Neural Process Network
- Search for Diffuse Supernova Neutrino Background in the Full KamLAND Dataset with Neural-Network-Based Event Classification
- Search for Diffuse Supernova Neutrino Background in the Full KamLAND Dataset with Neural-network-based Event Classification
BibTeX
@article{earl2026naptime,
title = {{NAPTIME: A Neural-process Framework for Rubin Alert Classification}},
author = {Nicholas Earl and Siddharth Chaini and K. Decker French and Jason T. Hinkle and Yashasvi Moon and Margaret Shepherd},
journal = {The Astrophysical Journal},
year = {2026},
month = oct,
doi = {10.3847/1538-4357/aea0d1},
url = {https://iopscience.iop.org/article/10.3847/1538-4357/aea0d1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - NAPTIME: A Neural-process Framework for Rubin Alert Classification
AU - Nicholas Earl
AU - Siddharth Chaini
AU - K. Decker French
AU - Jason T. Hinkle
AU - Yashasvi Moon
AU - Margaret Shepherd
PY - 2026
DA - 2026/10/01
JO - The Astrophysical Journal
DO - 10.3847/1538-4357/aea0d1
UR - https://iopscience.iop.org/article/10.3847/1538-4357/aea0d1
AB - The Vera C. Rubin Observatory Legacy Survey of Space and Time will produce a high-volume stream of irregularly sampled multiband alerts for which spectroscopic confirmation will be available only for a small minority of sources. Tidal disruption events are rare phenomena that provide a direct probe of dormant massive black holes, but their light curves can be confused with nuclear variability and other transient subclasses. We present NAPTIME (Neural Astrophysical Photometric Transient Identification and Modeling Engine), a neural-process framework for photometric transient classification under sparse and partial observational context. NAPTIME models irregular multiband light curves directly, combining probabilistic light-curve reconstruction with classification and optional host-galaxy context, as well as photometric-redshift information. We evaluate on two simulated benchmarks: ELAsTiCC2, our primary Rubin-like broad-classification benchmark, and MALLORN, a photometry-only tidal disruption event (TDE) focused benchmark. On the 15 family ELAsTiCC2 task, the metadata-aware model reaches macro F1 = 0.903 and macro area under the receiver operating characteristic curve (AUROC) = 0.991, while a matched photometry-only variant reaches 0.874 and 0.986. Viewed as a TDE-versus-rest ranking model, the classifier yields TDE average precision 0.985 with metadata and 0.979 without. Metadata is most valuable in the low-context regime. Using only the earliest 10% of retained photometric measurements, macro F1 is ∼0.42 with metadata and ∼0.34 without it. On MALLORN, NAPTIME reaches macro F1 = 0.693 and macro AUROC = 0.958. These results show that neural processes provide a practical probabilistic framework for Rubin-like transient classification and remain effective for TDE-focused candidate recovery.
SN - 0004-637X
SN - 1538-4357
ER -