原著者:Phil R. Van-Lane (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Dunlap Institute for Astronomy and Astrophysics, University of Toronto, Department of Astronomy and APhil R. Van-Lane (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Dunlap Institute for Astronomy and Astrophysics, University of Toronto, Department of Astronomy and Astrophysics, University of California San Diego), Joshua S. Speagle (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Department of Statistical Sciences, University of Toronto, Dunlap Institute for Astronomy and Astrophysics, University of Toronto, Data Sciences Institute, University of Toronto), Ryan Cloutier (Department of Physics and Astronomy, McMaster University), Christopher A. Theissen (Department of Astronomy and Astrophysics, University of California San Diego), Gwendolyn M. Eadie (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Department of Statistical Sciences, University of Toronto, Data Sciences Institute, University of Toronto), Ilay Kamai (Physics Department, Technion Israel Institute of Technology)
原著者: Phil R. Van-Lane (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Dunlap Institute for Astronomy and Astrophysics, University of Toronto, Department of Astronomy and Astrophysics, University of California San Diego), Joshua S. Speagle (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Department of Statistical Sciences, University of Toronto, Dunlap Institute for Astronomy and Astrophysics, University of Toronto, Data Sciences Institute, University of Toronto), Ryan Cloutier (Department of Physics and Astronomy, McMaster University), Christopher A. Theissen (Department of Astronomy and Astrophysics, University of California San Diego), Gwendolyn M. Eadie (David A. Dunlap Department of Astronomy and Astrophysics, University of Toronto, Department of Statistical Sciences, University of Toronto, Data Sciences Institute, University of Toronto), Ilay Kamai (Physics Department, Technion Israel Institute of Technology)
アーキテクチャ: EncoTESSは、他のTSFMで一般的な計算コストの高いTransformerアーキテクチャではなく、minGRU(Feng et al. 2024)に基づいたリカレントニューラルネットワーク(RNN)を利用している。この選択により、モデルは長いシーケンスを効率的に処理でき(O(L2)に対しO(L)のスケーリング)、約64,000個のパラメータのみで一般的なコンシューマ向けハードウェア上で動作可能である。
変動性の特性評価: 潜在空間(Θ)は、メタデータや短期的コンテキストのみを使用するモデルよりも高い精度で、伝統的な変動要約統計量(回転周期、フレア等価持続時間、歪度、尖度)を回収することに成功した。エンコーディングは、物理的特性(例:食連星 vs フレアを起こすM矮星)ごとに恒星をクラスタリングし、色、マグニチュード、および変動振幅と強く相関している。