Knot signature bounds from hyperbolic geometry

A geometric bound on knot signature

Model / AI: DeepMind neural networks

Davies and collaborators used supervised learning and feature attribution to identify a connection between knot signature and natural slope. The published result bounds their discrepancy using volume and injectivity radius; the mathematical proof was developed by humans.

Source read 2026-09-08. Located through seewoo5/awesome-ai-for-math. Read the Nature paper’s knot-theory results and methods. The paper is peer-reviewed; no computations or proofs were rerun.

• Paper — Nature paper
• Code — Code and datasets