Reproduce “Indirect Query Bayesian Optimization with Integrated Feedback” 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

Mengyan Zhang, Shahine Bouabid, Cheng Soon Ong, Seth Flaxman, Dino Sejdinovic

IQBO chooses indirect queries to optimize unknown function when observations average over possible inputs. CMES scores queries for information their feedback provides about the function’s maximum value, rather than the maximum average. In synthetic two-dimensional Branin experiments, CMES had lower simple and instant regret than adapted MES, UCB, and EI, especially early; its theorem bounds instant regret under stated assumptions.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.