Reproduce “Colombo: A Frontier-to-Frontier Autoresearch System for Oracle-Based Optimization” 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

Haihan Zhang, Wendao Wu, Chenheng Zhang, Chunyuan Zheng, Yanyi Li, Cong Fang, Haoxuan Li, Zhouchen Lin

Colombo is a research system for oracle-based optimization that maps gaps between achievable upper bounds and lower bounds, assigns adaptive research tasks, and challenges proposed algorithms with hard problem families. The authors report optimal or near-optimal rates in more than 50 Complexity Atlas settings. The paper does not compare Colombo with alternatives or isolate its components’ contributions.

Summary by alphaXiv.

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