Reproduce “Curriculum-Guided Acquisition of Hierarchical Task Knowledge from LLM-Generated Experience” 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

Yuesheng Xu, Hector Munoz-Avila

The authors turn verified LLM solutions to small planning problems into ordered state–task learning problems for an HTN learner, teaching component methods before larger tasks. They evaluate automated curricula against expert-crafted curricula and online learning across four domains. Automated curricula enable method reuse before testing, but upfront token costs make total use higher in the two simpler domains.

Summary by alphaXiv.

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