Reproduce “Markov Chain Truncation for Doubly-Intractable Inference” 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
Colin Wei, Iain Murray
The authors introduce coupled Markov-chain estimators for reciprocal normalizing constants that are intractable in parameter inference. Importance-sampling weights drive the chains, and randomized truncation debiases the reciprocal estimate; coupling lets later corrections cancel. In an Ising pseudo-marginal experiment, RBBCE produced 99,924 positive estimates in 100,000 iterations, versus 96,538 for IAE, when importance sampling was reliable; positivity is not guaranteed generally.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.