发表机构
University of Illinois Urbana-Champaign; Indian Institute of Technology Delhi; Yardi School of Artificial Intelligence (ScAI); National Center for Supercomputing Applications (NCSA)（伊利诺伊大学厄巴纳-香槟分校; 印度理工学院德里分校; 雅迪人工智能学院; 国家超级计算应用中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对异构科学接口，提出GEODE基础模型，耦合任务接口与小波算子库，联合预训练后仅训练2.1%参数即可迁移新任务，并区分了多任务覆盖、保持与预训练复用三个独立属性。
AI 中文摘要
科学基础模型通常在异构物理问题已被转换为兼容的网格化、标记化或符号化表示之后进行评估。这使得科学接口游离于预训练模型和实际复用审计之外。我们研究了互补的设置，其中边界历史、稀疏监测记录和载荷历史保留其原生推理类别，其输出仍位于笛卡尔、经纬度和非结构化域上。GEODE将任务特定的科学接口与共享的定向小波算子库耦合。一个联合预训练模型表示腔流、辐射剂量和弹塑性应力，然后通过训练包含其2.1%参数的私有接口获得热交换器和反应堆子通道。早期预测通过参数隔离保持不变，而无限制微调则使其退化14-29倍。关键在于，保持本身并不确立复用：范数匹配的随机化库对照表明，预训练计算的贡献取决于任务和数据机制。一项单独的分解表明，当场级主导范数时，全场相对L2误差可能大幅低估相对于空间变化的误差。任务特定算子在五个问题中的三个上仍然更准确。这些结果将多任务覆盖、保持和预训练复用区分为独立属性，当科学基础模型跨越异构接口时必须独立测试。
英文摘要
Scientific foundation models are commonly evaluated after heterogeneous physical problems have already been translated into a compatible gridded, tokenized or symbolic representation. This leaves the scientific interface outside both the pretrained model and the audit of what is actually reused. We study the complementary setting in which boundary histories, sparse monitor records and loading histories retain their native inference classes and their outputs remain on Cartesian, latitude-longitude and unstructured domains. GEODE couples task-specific scientific interfaces to a shared routed library of wavelet operators. A single jointly pretrained model represents cavity flow, radiation dose and elastoplastic stress, then acquires a heat exchanger and a reactor subchannel by training a private interface containing 2.1% of its parameters. Earlier predictions remain unchanged by parameter isolation, whereas unrestricted fine-tuning degrades them by factors of 14-29. Crucially, preservation alone does not establish reuse: norm-matched randomized-library controls show that the contribution of pretrained computation is conditional on the task and data regime. A separate decomposition shows that full-field relative L2 error can substantially understate error relative to spatial variation when field level dominates the norm. Task-specific operators remain more accurate on three of the five problems. These results distinguish multi-task coverage, preservation and pretrained reuse as separate properties that must be tested independently when scientific foundation models span heterogeneous interfaces.
Comments71 pages, 6 figures, 18 supplementary figures
