LGCOJan 30, 2025

Neural Discovery in Mathematics: Do Machines Dream of Colored Planes?

arXiv:2501.18527v34 citationsh-index: 29ICML
Originality Incremental advance
AI Analysis

This work addresses a long-standing open problem in mathematics, offering a new computational approach for geometric discovery, though it is incremental in advancing a specific variant.

The paper tackled the Hadwiger-Nelson problem in discrete geometry by reformulating it as an optimization task using neural networks, resulting in the discovery of two novel six-colorings that improved upon a variant of the problem for the first time in thirty years.

We demonstrate how neural networks can drive mathematical discovery through a case study of the Hadwiger-Nelson problem, a long-standing open problem at the intersection of discrete geometry and extremal combinatorics that is concerned with coloring the plane while avoiding monochromatic unit-distance pairs. Using neural networks as approximators, we reformulate this mixed discrete-continuous geometric coloring problem with hard constraints as an optimization task with a probabilistic, differentiable loss function. This enables gradient-based exploration of admissible configurations that most significantly led to the discovery of two novel six-colorings, providing the first improvement in thirty years to the off-diagonal variant of the original problem. Here, we establish the underlying machine learning approach used to obtain these results and demonstrate its broader applicability through additional numerical insights.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes