发表机构
Stony Brook University; Westlake University（石溪大学; 西湖大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对科学动力学预测，提出协议相关模型选择框架ProtocolMatch，通过计算匹配与验证选择评估多种预测器，发现模型排名随协议变化，强调按协议报告准确性、物理有效性与分布可靠性。
AI 中文摘要
科学动力学预测常被视为架构选择问题，然而实际部署还取决于观测历史、滚动反馈、计算预算、物理目标以及测试分布。我们提出了协议相关模型选择问题，并引入ProtocolMatch，一个计算匹配、验证选择且保留失败案例的评估框架。在受驱动的量子自旋动力学中，我们比较了循环网络、分块注意力、因果注意力和低秩线性预测器在三个独立生成数据集上的表现。在固定的双自旋任务中，随着训练集增大，因果注意力与循环网络的排序发生反转；而在六自旋局部可观测量比较中，线性预测器具有最低的平均误差。在四自旋研究中，限制观测历史会恶化所有刷新历史视图，但改善所有闭环视图。在状态刷新下，最新状态MLP在所有五个单元中的每个数据集上误差均低于持久性预测，但其闭环排名因系统而异，并包含有限爆炸误差。物理惩罚能改善目标一致性，但未能可靠地改善预测误差；在驱动频率偏移后，分布内区间大多失去覆盖。因此，科学模型选择应返回预测器及其协议，并分别报告准确性、物理有效性和偏移分布可靠性。
英文摘要
Scientific dynamics forecasting is often framed as an architecture choice, although deployment is also determined by observed history, rollout feedback, compute budget, physical objective, and test distribution. We formulate protocol-dependent model selection and introduce ProtocolMatch, a compute-matched, validation-selected, and failure-preserving evaluation framework. On driven quantum-spin dynamics, we compare recurrent, patched-attention, causal-attention, and low-rank linear predictors across three independently generated datasets. The causal-attention--recurrence ordering reverses as the training set grows within a fixed two-spin task, while a linear predictor has the lowest mean error in the six-spin local-observable comparison. Restricting observed history worsens every refreshed-history view but improves every closed-loop view in the four-spin study. A latest-state MLP has lower error than persistence on every dataset under state refresh across all five cells, yet its closed-loop rank varies by system and includes finite explosive errors. Physical penalties improve targeted consistency without reliably improving prediction error, and in-distribution intervals lose most coverage after a driving-frequency shift. Thus scientific model selection should return a predictor with its protocol and report accuracy, physical validity, and shifted-distribution reliability separately.
Comments14 pages, 4 figures, 8 tables
