Reproduce “Improved machine learning algorithm for predicting ground state properties” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Laura Lewis, Hsin-Yuan Huang, Viet T. Tran, Sebastian Lehner, Richard Kueng, John Preskill
Researchers from Caltech and Johannes Kepler University developed an improved classical machine learning algorithm for predicting ground state properties of quantum many-body systems. The algorithm reduces training data requirements to O(log n) from prior O(n^c) and achieves quasi-polynomial scaling in prediction error, while demonstrating nearly linear computational time in system size.
Key insights
- Explicitly encoding geometric locality as an inductive bias within classical ML models significantly improves their efficiency for quantum many-body problems.
- Ground state properties of gapped local Hamiltonians can be accurately represented as a sum of local, smooth functions, enabling efficient approximation in feature space.
- A novel ℓ_1-norm bound for Pauli coefficients is crucial for theoretically proving the favorable scaling of ℓ_1-regularized regression in this specific quantum context.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.