GraphCast中大气河的机制可解释性
Mechanistic Interpretability of Atmospheric Rivers in GraphCast
浏览论文内容
中文总结 AI 辅助
本研究在GraphCast上训练稀疏自编码器，发现其内部稳定计算大气河强度（IVT），并验证了因果性，为理解AI天气模型内部机制提供了新方法。
中文摘要 AI 辅助
虽然人工智能天气模型现已可与业务预报相媲美，但它们如何在内部表示大气仍是一个悬而未决的问题：特征归因揭示了哪些输入模式重要，而非模型计算了什么或如何在内部组合信息。我们在GraphCast上训练稀疏自编码器（SAEs）以揭示其学到的概念，并以大气河作为我们关注的现象。标准和Matryoshka SAEs均显示，GraphCast将大气河强度（以积分水汽输送（IVT）衡量）计算为一个稳定的内部变量，尽管IVT既非输入也非目标。与标准SAE的无结构概念检索相比，Matryoshka SAE按重要性对概念进行排序并揭示它们之间的关系。大气河概念在深度上持续存在，直接干预确认了因果关系。该方法提供了一种找到内部变量并确定模型实际依赖其中哪些变量的途径，这是询问在气候变暖下现象变化时这些变量是否仍具意义的前提。
英文摘要
While AI weather models now rival operational forecasts, how they represent the atmosphere internally remains an open question: feature attribution reveals which input patterns matter, not what the model computes or how it combines information internally. We train sparse autoencoders (SAEs) on GraphCast to uncover its learned concepts, using atmospheric rivers as our phenomenon of focus. Both standard and Matryoshka SAEs show GraphCast computes atmospheric river intensity, measured by integrated vapor transport (IVT), as a stable internal variable, despite IVT being neither an input nor a target. In contrast to the unstructured concept retrieval of the standard SAE, the Matryoshka SAE orders concepts by importance and exposes their relations. Atmospheric river concepts persist across depth and direct interventions confirm causality. This method offers a way to find internal variables and determine which of them the model actually relies on, which is a prerequisite for asking whether those variables remain meaningful as the phenomenon changes under a warming climate.
发表机构
- University of Virginia（弗吉尼亚大学）
机构由 AI 辅助整理，请以论文原文为准。
