arXiv 2510.237492025-10astro-ph.IMcs.LG

重新构想欧几里得星系形态学：利用稀疏自编码器识别和解释特征

Re-envisioning Euclid Galaxy Morphology: Identifying and Interpreting Features with Sparse Autoencoders

Space Telescope Science Institute（太空望远镜科学研究所）；Johns Hopkins University（约翰霍普金斯大学）；University of Toronto（多伦多大学）

John F. Wu, Michael Walmsley

备注：: Authors contributed equally to this work. Accepted to NeurIPS Machine Learning and the Physical Sciences Workshop. See trained model at https://huggingface.co/mwalmsley/euclid-rr2-mae, HuggingFace demo at https://huggingface.co/spaces/mwalmsley/euclid_masked_autoencoder, and code at https://github.com/jwuphysics/euclid-galaxy-morphology-saes

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

该研究利用稀疏自编码器（SAEs）从预训练神经网络中识别欧几里得星系形态的单语义特征，通过监督模型和自监督模型验证，发现SAEs能识别星系动物园决策树外的可解释特征，且与标签对齐度优于PCA，为天体物理现象发现提供工具。

摘要：

稀疏自编码器（SAEs）可从预训练神经网络中高效识别星系形态的候选单语义特征。我们在欧几里得Q1图像上使用监督模型（Zoobot）和新的自监督模型（MAE）进行了验证。我们公开的MAE实现了超人类的图像重建性能。监督模型上的主成分分析（PCA）主要识别与星系动物园决策树对齐的特征，而SAEs能识别该框架外的可解释特征。SAE特征与星系动物园标签的对齐度也强于PCA。尽管可解释性仍存挑战，但SAEs为发现人类定义分类之外的天体物理现象提供了强大工具。

英文摘要：

Sparse Autoencoders (SAEs) can efficiently identify candidate monosemantic features from pretrained neural networks for galaxy morphology. We demonstrate this on Euclid Q1 images using both supervised (Zoobot) and new self-supervised (MAE) models. Our publicly released MAE achieves superhuman image reconstruction performance. While a Principal Component Analysis (PCA) on the supervised model primarily identifies features already aligned with the Galaxy Zoo decision tree, SAEs can identify interpretable features outside of this framework. SAE features also show stronger alignment than PCA with Galaxy Zoo labels. Although challenges in interpretability remain, SAEs provide a powerful engine for discovering astrophysical phenomena beyond the confines of human-defined classification.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。