Reproduce “Sample-efficient learning with Dense Associative Memories” 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
Andrew Lucas, Dmitry Krotov
The authors test whether a Dense Associative Memory can learn parity rules rather than memorize valid bitstrings. The method thresholds weak correlations, retaining persistent parity-check signals. For uniformly sampled codewords of bounded-check codes with distance greater than one, a positive threshold gives a sufficient O(log n) sample bound for recovering codewords as ground states, not for arbitrary data.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.