Reproduce “Information, Estimation, and Prediction for Discrete Memoryless Channels via Poissonization” 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
Abhiram Gorle, Tsachy Weissman
The authors study information about a hidden input and prediction of outputs from a noisy channel. They Poissonize observation count with mean γ and show information gain per unit budget equals expected KL divergence between the true channel row and posterior predictive distribution. Seven coupon-collector triples retain 95.3% of full-input information at one expected read, but are not capacity-optimal at any positive budget.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.