SELDON: Supernova Explosions Learned by Deep ODE Networks

2026-03-04

AI Summary

This work addresses the challenge of real-time processing of tens of millions of nightly optical transient alerts from the Rubin Observatory, a task beyond the capacity of traditional physics-based pipelines. The authors propose a continuous-time variational autoencoder that uniquely integrates a masked GRU-ODE encoder, a latent-space neural ODE propagator, and an interpretable Gaussian mixture decoder to effectively model sparse, irregularly sampled, non-stationary, and heteroscedastic multi-band supernova light curves. The method enables continuous-time extrapolation under extremely sparse observational data and achieves millisecond-level accurate estimation of key physical parameters—such as rise time, decay rate, and peak flux—thereby significantly enhancing the efficiency of prioritizing spectroscopic follow-up observations in large-scale time-domain surveys.

Technology Category

• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Mixed Discrete/Continuous Search
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs

Abstract

The discovery rate of optical transients will explode to 10 million public alerts per night once the Vera C. Rubin Observatory's Legacy Survey of Space and Time comes online, overwhelming the traditional physics-based inference pipelines. A continuous-time forecasting AI model is of interest because it can deliver millisecond-scale inference for thousands of objects per day, whereas legacy MCMC codes need hours per object. In this paper, we propose SELDON, a new continuous-time variational autoencoder for panels of sparse and irregularly time-sampled (gappy) astrophysical light curves that are nonstationary, heteroscedastic, and inherently dependent. SELDON combines a masked GRU-ODE encoder with a latent neural ODE propagator and an interpretable Gaussian-basis decoder. The encoder learns to summarize panels of imbalanced and correlated data even when only a handful of points are observed. The neural ODE then integrates this hidden state forward in continuous time, extrapolating to future unseen epochs. This extrapolated time series is further encoded by deep sets to a latent distribution that is decoded to a weighted sum of Gaussian basis functions, the parameters of which are physically meaningful. Such parameters (e.g., rise time, decay rate, peak flux) directly drive downstream prioritization of spectroscopic follow-up for astrophysical surveys. Beyond astronomy, the architecture of SELDON offers a generic recipe for interpretable and continuous-time sequence modeling in any time domain where data are multivariate, sparse, heteroscedastic, and irregularly spaced.

Problem

• supernova
• light curves
• time-series forecasting
• sparse data
• irregular sampling

Innovation

• neural ODE
• continuous-time modeling
• interpretable deep learning
• irregular time series
• variational autoencoder