基于基础模型摘要的跨模拟器迁移：迈向SKA时代的稳健再电离推断
Cross-simulator transfer with foundation model summaries: Towards robust SKA-era reionization inference
查看机构详情
- Institut für Theoretische Physik, Universität Heidelberg（海德堡大学理论物理研究所）
- Max Planck Institute for Radioastronomy（马克斯·普朗克射电天文研究所）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
该研究提出用在半数值模拟上预训练的基础模型SKATR，实现跨模拟器迁移的再电离推断，精度达监督基线水平且模拟量减少2.6倍，在SKA噪声下仍表现优异。
中文摘要 AI 辅助
基于模拟的参数推断（SBI）易受模型误设影响：在特定正向模型上训练的神经摘要和密度估计器，应用于另一模型生成的数据或真实观测数据时通常失效，且没有任何训练模拟器能完全捕捉真实测量的完整观测流程。本文展示，在快速近似模拟器上进行无标签预训练的自监督视觉Transformer（ViT）可生成可跨模拟器泛化的可迁移数据摘要，无需重新训练即可作为冻结编码器，用于从完全不同的、显式解析辐射传输的模拟器数据中推断天体物理参数，而该编码器从未见过该模拟器的数据或参数。作为21cm宇宙学的具体应用案例，SKATR（一种用联合嵌入预测架构（JEPA）预训练的ViT）是用于从即将到来的SKA测量中进行再电离推断的基础模型：SKATR先在6.7万个低成本、无噪声的半数值21cmFAST光锥上进行一次预训练，随后被冻结并应用于流体动力学Loreli II光锥，其中一个轻量条件流匹配头用于推断5个天体物理参数；编码器从未见过Loreli数据、其参数或任何噪声。对比实验显示，SKATR在所有5个参数上都产生最精确、校准最佳的后验分布，达到完全监督域内基线的精度，同时所需的辐射传输模拟量减少2.6倍。在真实SKA AA*噪声下，仅SKATR同时保持准确、信息丰富且校准良好，甚至优于从噪声数据从头重新训练的监督基线。因此，在计算高效的半数值模拟上进行自监督预训练，是实现SKA时代校准良好、模拟器无关且噪声无关的再电离推断的可行路径。
英文摘要
Simulation-based inference (SBI) for parameter estimation is vulnerable to model misspecification: neural summaries and density estimators trained on a specific forward model typically fail when applied to data drawn from another model, or from real observations, and no training simulator can capture the full observational pipeline of a real measurement exactly. We show that a self-supervised Vision Transformer (ViT), pretrained label-free on a fast approximate simulator, produces transferable data summaries that generalize across simulators. Without retraining, it can be reused as a frozen encoder to infer astrophysical parameters from a completely different simulator that resolves the radiative transfer explicitly, on which it has never seen either data or parameters. As a concrete use case in 21cm cosmology, SKATR, a ViT pretrained with a Joint Embedding Predictive Architecture (JEPA), serves as a foundation model for reionization inference from upcoming SKA measurements: SKATR is pretrained once on 67k low-cost, noiseless semi-numerical 21cmFAST lightcones, then frozen and applied to hydrodynamical Loreli II lightcones, where a lightweight conditional flow matching head infers five astrophysical parameters; the encoder is never shown Loreli data, its parameters, or any noise. In our comparison, SKATR yields the most precise and best-calibrated posteriors across all five parameters, matching the accuracy of the fully-supervised in-domain baseline while requiring 2.6x fewer radiative-transfer simulations. Under realistic SKA AA* noise, only SKATR remains simultaneously accurate, informative, and calibrated, outperforming even a supervised baseline retrained from scratch on noisy data. Self-supervised pretraining on computationally efficient semi-numerical simulations is therefore a viable route to calibrated, simulator- and noise-agnostic reionization inference for the SKA-era.
