Astroconformer: Inferring Surface Gravity of Stars from Stellar Light Curves with Transformer

TL;DR

Astroconformer is a Transformer-based model that accurately infers stellar surface gravity from light curves, outperforming previous methods and providing interpretable insights into stellar oscillations, applicable to both space and ground-based data.

Contribution

The paper introduces Astroconformer, the first Transformer model for stellar surface gravity inference, demonstrating superior performance and interpretability over existing data-driven approaches.

Findings

Outperforms state-of-the-art methods in surface gravity inference

Utilizes long-range attention to capture stellar oscillation patterns

Generalizes well to sparse ground-based light curves

Abstract

We introduce Astroconformer, a Transformer-based model to analyze stellar light curves from the Kepler mission. We demonstrate that Astrconformer can robustly infer the stellar surface gravity as a supervised task. Importantly, as Transformer captures long-range information in the time series, it outperforms the state-of-the-art data-driven method in the field, and the critical role of self-attention is proved through ablation experiments. Furthermore, the attention map from Astroconformer exemplifies the long-range correlation information learned by the model, leading to a more interpretable deep learning approach for asteroseismology. Besides data from Kepler, we also show that the method can generalize to sparse cadence light curves from the Rubin Observatory, paving the way for the new era of asteroseismology, harnessing information from long-cadence ground-based observations.