Abstract:In 2021, Adam Zsolt Wagner proposed an approach to disprove conjectures in graph theory using Reinforcement Learning (RL). Wagner's idea can be framed as follows: consider a conjecture, such as a certain quantity f(G) < 0 for every graph G; one can then play a single-player graph-building game, where at each turn the player decides whether to add an edge or not. The game ends when all edges have been considered, resulting in a certain graph G_T, and f(G_T) is the final score of the game; RL is then used to maximize this score. This brilliant idea is as simple as innovative, and it lends itself to systematic generalization. Several different single-player graph-building games can be employed, along with various RL algorithms. Moreover, RL maximizes the cumulative reward, allowing for step-by-step rewards instead of a single final score, provided the final cumulative reward represents the quantity of interest f(G_T). In this paper, we discuss these and various other choices that can be significant in Wagner's framework. As a contribution to this systematization, we present four distinct single-player graph-building games. Each game employs both a step-by-step reward system and a single final score. We also propose a principled approach to select the most suitable neural network architecture for any given conjecture, and introduce a new dataset of graphs labeled with their Laplacian spectra. Furthermore, we provide a counterexample for a conjecture regarding the sum of the matching number and the spectral radius, which is simpler than the example provided in Wagner's original paper. The games have been implemented as environments in the Gymnasium framework, and along with the dataset, are available as open-source supplementary materials.

Reinforcement learning for graph theory, II. Small Ramsey numbers

Reinforcement learning for graph theory, I. Reimplementation of Wagner's approach

Restricted size Ramsey numbers and restricted size Ramsey minimal graphs for matching versus wheel upto order six

RamseyRL: A Framework for Intelligent Ramsey Number Counterexample Searching

New bounds for Ramsey numbers $R(K_k-e,K_l-e)$

Three Results on Cycle-Wheel Ramsey Numbers

A Systematization of the Wagner Framework: Graph Theory Conjectures and Reinforcement Learning

Ramsey problems for monotone paths in graphs and hypergraphs

Small Ramsey numbers for books, wheels, and generalizations

Ramsey theory and strength of graphs

Lower bounds for Ramsey numbers as a statistical physics problem

Lower Bounds for Book Ramsey Numbers

Canonical Ramsey numbers of sparse graphs

On Ramsey Size-Linear Graphs and Related Questions

A canonical Ramsey theorem with list constraints in random graphs

Ramsey Achievement Games on Graphs : Algorithms and Bounds

Extending Ramsey Numbers for Connected Graphs of Size 3

Size Ramsey numbers of small graphs versus fans or paths

Some Open Problems of Ramsey Minimal Graphs

New Ramsey Multiplicity Bounds and Search Heuristics

Small minimal $(3, 3)$-Ramsey graphs