Scaling Sudoku as a Constraint Problem
← All talks

Talk ·

Scaling Sudoku as a Constraint Problem

What changes when Sudoku grows from 6×6 to 36×36, based on propagation experiments and a public corpus of 434,201 generated instances.

Speaker
Mikael Zayenz Lagerkvist

Interactive presentation

Slides

Open slides in a new view

Click the presentation, then use the arrow keys, Page Up/Page Down, or Space to move through it.

This talk revisits Sudoku as a constraint-programming benchmark at five grid sizes. It presents the 434,201-instance corpus, explains the propagation-based hardness categories used to map it, and shows how sharply the distribution changes as the grid grows.