Reproduce “Standards-Guided and Dependency-Aware Code-Level Documentation Generation” 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
Tingxu Han, Wei Song, Ziming Li, Yuchen Chen, Weisong Sun, Zhenpeng Chen, Chunrong Fang, Zhenyu Chen and 1 more
CODEDOC generates entity documentation using selected caller and callee source, repository paths, and standards-derived criteria. It applies those criteria to judge generated documentation and guide one refinement pass. Across ten open-source Python repositories, it scores 3.79 under ISO multi-round evaluation, versus 3.24 for RepoAgent and 2.93 for Long-Context, using roughly an order of magnitude fewer tokens per entity than Long-Context.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.