English(EN) PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates

PhysGuard框架改进神经算子从仿真到现实的适配

作者 PulseAugur 编辑部 · [2 个来源] · 2026-06-15 11:50

研究人员开发了PhysGuard，一个旨在改进神经算子从仿真到现实适配的新框架。该方法利用仿真数据中的Fisher信息矩阵，在微调过程中识别并保护物理关键参数方向。PhysGuard旨在防止在标准微调过程中可能发生的、尤其是在显著的领域偏移下，基本物理表示的退化。实验表明，与传统的微调方法相比，PhysGuard可将低频误差最多降低32%，同时保持了适应性。 AI

影响 PhysGuard为弥合神经算子的仿真到现实差距提供了一种新颖的方法，有可能提高科学模拟中使用的模型的准确性和可靠性。

排序理由该集群包含一篇发表在arXiv上的研究论文，详细介绍了一种新的神经算子适配方法。

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.LG TIER_1 English(EN) · Changjian Zhou, Junfeng Fang, Negin Yousefpour, Peng Wu, Bin Yan, Guillermo A Narsilio · 2026-06-16 04:00

PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates

arXiv:2606.16602v1 Announce Type: new Abstract: Neural operator models trained on simulation data often lose accuracy when applied to experimental measurements due to the sim-to-real gap. Standard fine-tuning with limited real data can reduce this gap, but it may also damage the …
arXiv cs.LG TIER_1 English(EN) · Guillermo A Narsilio · 2026-06-15 11:50

PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates

Neural operator models trained on simulation data often lose accuracy when applied to experimental measurements due to the sim-to-real gap. Standard fine-tuning with limited real data can reduce this gap, but it may also damage the core physics-relevant representations learned du…

报道来源 [2]

PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates

PhysGuard: Fisher-Guided Gradient Projection for Sim-to-Real Neural PDE Surrogates

相关实体

相关话题