English(EN) Stability-Constrained Approximation in Spline KANs: Exact Layer Balancing and Budget-Compatible Saturation

关于样条 KANs 的新理论解决了近似-稳定性张力

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-17 04:00

本文介绍了一种在严格的逐层 Lipschitz 预算内近似深度样条叠加网络的新方法。它定义了两个关键量：因子稳定性复杂度和预算兼容近似复杂度。该研究为有限深度对角线平衡问题提供了精确解，并提出了一个保持预算且具有受控松弛的构造性样条离散化定理。此外，它还建立了线性样条值算子的 minimax 下界，并证明了层误差在组合下不一定会抵消。 AI

影响引入了理论框架，用于在计算约束下提高深度样条网络的稳定性和近似能力。

排序理由该项目是一篇学术论文，详细介绍了样条 KANs 的理论进展。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv stat.ML 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 17 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该项目是一篇学术论文，详细介绍了样条 KANs 的理论进展。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv stat.ML TIER_1 English(EN) · Aleksander Tankman · 2026-09-17 04:00

Spline KANs 中的约束稳定性近似：精确层平衡与预算兼容饱和

arXiv:2609.17619v1 Announce Type: new Abstract: Deep spline superposition networks face a tension between approximation order and stability across depth. We study approximation under a hard layerwise Lipschitz budget, and organise it around two quantities: the factorisation stabi…

本文如何被排名

报道来源 [1]

Spline KANs 中的约束稳定性近似：精确层平衡与预算兼容饱和

相关话题