Researchgraph neural networksquantum entanglementsdrgfinite temperature
GNN Learns SDRG Decimation For Quantum Entanglement
6.1
Relevance Score
On March 5, 2026 Stefan Kettemann posted an arXiv preprint training machine learning to infer entanglement in disordered long-range S=1/2 quantum spin chains using the strong disorder renormalisation group (SDRG) as a teacher. A graph neural network achieves near-perfect pairing accuracy, reproduces entanglement entropy S(ℓ) across interaction exponents, and extends to finite-temperature properties via an SDRGX two-stage strategy without retraining.

