面向科学 AI 的可观测特定有效性测试：受控近似与干预响应
Observable-Specific Validity Tests for Scientific AI: Controlled Approximations and Intervention Response
AI 导读
arXiv 论文提出面向科学 AI 的“可观测特定有效性记录”框架，将近似与其参考模型、误差度量、容差、参数域及干预变换绑定。实验中，在 16 次重复神经拟合中加入量纲齐次约束后，能量尺度迁移误差从 52.3% 降至 0.0768%，同时满足能隙与响应容差的中位占比从零升至 100%；论文还推导出随 $x^{-2k}$ 缩放的固定控制查询需求，并配套 256 任务评测接口。
正文 · 原文
Abstract:An approximation can reproduce an energy while missing its response, or satisfy an error tolerance without being certified by a given bound. We formulate observable-specific validity records for scientific artificial intelligence (AI) that link a reduction to its reference model, error measure, tolerance, parameter domain, and transformations or interventions. Explicit displacement and velocity certificates for a singularly perturbed oscillator recover distinct convergence orders and provide sufficient validity guarantees. Exact Hubbard dimer relations resolve the different coupling domains accepted by energy and response criteria. Across 16 repeated neural fits, imposing dimensional homogeneity reduces energy scale transfer error from 52.3% to 0.0768% and raises the median fraction jointly satisfying gap and response tolerances from zero to 100%. Paired transformations verify the protected scaling relation, while a coupling crossover identifies the additional information needed for response transfer. A causal model recovers a positive intervention response beneath a strong negative observational association. For two mechanisms observed with Gaussian noise whose responses first differ at order $x^k$, we derive an exact fixed control query requirement scaling as $x^{-2k}$ and verify its full success curve numerically. Together with a 256-task evaluation interface, these results provide independently checkable tests of controlled approximation, physical transfer, and mechanism discrimination in scientific AI.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Digital Libraries (cs.DL) |
| Cite as: | arXiv:2610.41939 [cs.LG] |
| (or arXiv:2610.41939v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41939 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Xiao-Sheng Ni [view email]
[v1]
Fri, 30 Oct 2026 15:57:51 UTC (294 KB)
来源：arXiv cs.AI · arxiv.org