Variational diffusion transformers for conditional sampling of supernovae spectra

2025-05-05

AI Summary

To address systematic uncertainties in cosmological analyses arising from high spectral diversity and the lack of physically grounded models for Type Ia supernovae, this work proposes the first cross-modal spectral generation framework integrating variational diffusion with a conditional Transformer, conditioned on light curves to reconstruct high-fidelity spectra across all phases. The method leverages radiative-transfer simulations to construct a synthetic dataset and learns the mapping from light curves to spectra. Experiments demonstrate substantial improvements: the full-phase mean squared error (MSE) drops to 0.108—79% lower than SALT3—with post-peak MSE reduced to 0.0191, just one-tenth that of SALT3. Moreover, the coverage of reconstructed spectral credible intervals closely matches nominal levels. This framework significantly enhances the identification of anomalous supernovae under sparse sampling and improves the accuracy, robustness, and physical interpretability of spectral inference.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Computer Vision: Diffusion Models for Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Type Ia Supernovae (SNe Ia) have become the most precise distance indicators in astrophysics due to their incredible observational homogeneity. Increasing discovery rates, however, have revealed multiple sub-populations with spectroscopic properties that are both diverse and difficult to interpret using existing physical models. These peculiar events are hard to identify from sparsely sampled observations and can introduce systematics in cosmological analyses if not flagged early; they are also of broader importance for building a cohesive understanding of thermonuclear explosions. In this work, we introduce DiTSNe-Ia, a variational diffusion-based generative model conditioned on light curve observations and trained to reproduce the observed spectral diversity of SNe Ia. In experiments with realistic light curves and spectra from radiative transfer simulations, DiTSNe-Ia achieves significantly more accurate reconstructions than the widely used SALT3 templates across a broad range of observation phases (from 10 days before peak light to 30 days after it). DiTSNe-Ia yields a mean squared error of 0.108 across all phases-five times lower than SALT3's 0.508-and an after-peak error of just 0.0191, an order of magnitude smaller than SALT3's 0.305. Additionally, our model produces well-calibrated credible intervals with near-nominal coverage, particularly at post-peak phases. DiTSNe-Ia is a powerful tool for rapidly inferring the spectral properties of SNe Ia and other transient astrophysical phenomena for which a physical description does not yet exist.

Problem

• Modeling diverse Type Ia Supernovae spectra from sparse observations
• Improving spectral reconstruction accuracy over existing templates
• Enabling spectral inference without complete physical models

Innovation

• Variational diffusion transformers for spectral sampling
• Conditional generative model for supernovae spectra
• Improved accuracy over SALT3 templates