Reproduce “Classical Hardness of Learning Functions of Hamiltonians” 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
Sota Hashimoto, Akinori Kawachi
The authors prove that accurate classical learning of two specified Hamiltonian function-learning problems would enable factoring random RSA moduli. The problems use fixed cosine or exponential functions and constructed distributions of succinct non-local projector Hamiltonians. A reduction compares predictions on paired inputs to recover factor bits, succeeding with probability at least 99/100 under the loss guarantee, assuming average-case factoring hardness.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.