VADER: A Variational Autoencoder to Infer Planetary Masses and Gas-Dust Disk Properties Around Young Stars

2025-09-15

AI Summary

Simultaneous inference of planetary mass and disk physical parameters—such as α-viscosity, dust-to-gas ratio, and Stokes number—from protoplanetary disk (PPD) observations remains challenging due to strong degeneracies and high-dimensional parameter coupling. Method: We propose the first probabilistic deep learning framework for PPD parameter inference, built upon a variational autoencoder (VAE), enabling end-to-end, uncertainty-aware mapping from high-resolution ALMA dust continuum images to multiple physical parameters. Trained on >100,000 synthetic images generated by FARGO3D+RADMC3D, the model jointly estimates planetary mass and global disk properties. Results: Applied to real ALMA data, it achieves R² > 0.9 and yields planetary mass estimates consistent with literature values. This work pioneers the use of probabilistic deep learning for planet–disk joint inversion, uncovering latent physical correlations in observational data and substantially enhancing the efficiency and reliability of PPD physical interpretation in the era of interferometric surveys.

Technology Category

• Machine Learning: Probabilistic Circuits and Graphical Models
• Reasoning under Uncertainty: Probabilistic Inference
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

We present extbf{VADER} (Variational Autoencoder for Disks Embedded with Rings), for inferring both planet mass and global disk properties from high-resolution ALMA dust continuum images of protoplanetary disks (PPDs). VADER, a probabilistic deep learning model, enables uncertainty-aware inference of planet masses, $α$-viscosity, dust-to-gas ratio, Stokes number, flaring index, and the number of planets directly from protoplanetary disk images. VADER is trained on over 100{,}000 synthetic images of PPDs generated from exttt{FARGO3D} simulations post-processed with exttt{RADMC3D}. Our trained model predicts physical planet and disk parameters with $R^2 > 0.9$ from dust continuum images of PPDs. Applied to 23 real disks, VADER's mass estimates are consistent with literature values and reveal latent correlations that reflect known disk physics. Our results establish VAE-based generative models as robust tools for probabilistic astrophysical inference, with direct applications to interpreting protoplanetary disk substructures in the era of large interferometric surveys.

Problem

• Infer planet masses from protoplanetary disk images
• Determine global disk properties like viscosity and dust ratio
• Enable uncertainty-aware inference directly from ALMA observations

Innovation

• Variational Autoencoder for disk parameter inference
• Trained on synthetic FARGO3D and RADMC3D images
• Directly predicts planet masses and disk properties