跳转至

面向格罗滕迪克常数的长周期AI数学研究：人机数学协作案例研究

文章背景与核心概要

本文详细探讨了长周期人工智能研究系统在推进格罗滕迪克常数（Grothendieck constant, \(K_G\)）边界方面的应用。格罗滕迪克常数是一个基础常数，用于衡量组合优化问题与其连续松弛之间的难度差距。通过利用能够生成真正新颖数学见解的AI系统，研究人员成功将该常数的已知最佳上下界收紧至：

\[\frac{6\pi}{11} \;\le\; K_G \;\le\; \frac{\pi}{2\log(1+\sqrt{2})} - 10^{-4}\]

该研究深入探讨了在实现突破性数学发现的过程中，促进高效人机协作所需的优势、劣势以及必要条件，为AI在纯数学及理论计算机科学中的应用提供了宝贵的实证经验。

摘要

Summary

This paper presents a detailed case study on the application of long-horizon AI research systems to advance bounds on the Grothendieck constant (\(K_G\)), a fundamental value capturing the hardness between combinatorial problems and their continuous relaxations. Utilizing an AI system capable of generating genuinely novel mathematical insights, researchers successfully tightened the best-known bounds to:

\[\frac{6\pi}{11} \;\le\; K_G \;\le\; \frac{\pi}{2\log(1+\sqrt{2})} - 10^{-4}\]

The study discusses the strengths, weaknesses, and necessary conditions for fostering effective human-AI collaboration to achieve breakthrough mathematical discoveries.

本文详细探讨了应用长周期AI研究系统来推进格罗滕迪克常数（\(K_G\)）边界的案例研究。格罗滕迪克常数是一个基础数值，用于捕捉组合问题与其连续松弛之间的难度。利用能够产生真正新颖数学见解的AI系统，研究人员成功将已知的最佳边界收紧至：

\[\frac{6\pi}{11} \;\le\; K_G \;\le\; \frac{\pi}{2\log(1+\sqrt{2})} - 10^{-4}\]

该研究探讨了促进有效人机协作以实现突破性数学发现的优势、劣势和必要条件。

文档元数据

Document Metadata

  • arXiv ID: arXiv:2608.11195 [cs.AI]
  • 主题分类（Subjects）: 人工智能 (cs.AI)；计算复杂性 (cs.CC)；人机交互 (cs.HC)；泛函分析 (math.FA)
  • 发布日期（Publication Dates）:
  • 提交于：2026年8月11日
  • 最后修订：2026年8月14日 (v3)
  • DOI: 10.48550/arXiv.2608.11195

作者

Authors

  • Alan Li
  • Rahul Saha
  • Anton Xue
  • Swarat Chaudhuri
  • Adam Klivans
  • Pravesh K Kothari
  • Raghu Meka

摘要原文

Abstract

AI agents are increasingly used in mathematics research, but it is often unclear how to use them effectively. Towards this, we present an extensive case study of how AI was used to improve bounds on the Grothendieck constant \(K_G\), which captures the hardness between combinatorial problems and their continuous relaxations. Specifically, while the precise value of \(K_G\) is not known, we recently tightened the best known bounds to:

\[\frac{6\pi}{11} \;\le\; K_G \;\le\; \frac{\pi}{2\log(1+\sqrt{2})} - 10^{-4}\]

Crucially, these improvements were achieved using an AI research system that could arrive at insights deemed novel by domain experts. We give a detailed discussion of our experience using AI for mathematics research, particularly touching upon its strengths and weaknesses, as well as our experience with creating ideal conditions for AI to arrive at breakthrough insights.

AI智能体在数学研究中的应用日益广泛，但如何有效地使用它们往往尚不明确。为此，我们深入剖析了一个案例研究，展示了如何利用AI来改善格罗滕迪克常数 \(K_G\) 的边界。\(K_G\) 衡量了组合问题与其连续松弛之间的难度。具体而言，尽管 \(K_G\) 的精确值尚不清楚，但我们最近将已知的最佳边界收紧至：

\[\frac{6\pi}{11} \;\le\; K_G \;\le\; \frac{\pi}{2\log(1+\sqrt{2})} - 10^{-4}\]

至关重要的是，这些进展是通过一个AI研究系统实现的，该系统能够得出被领域专家认为具有新颖性的见解。我们详细讨论了使用AI进行数学研究的经验，特别是其优势与劣势，以及为AI创造产生突破性见解的理想条件的经验。

资源与链接