Reproduce “LM-DSB: A Lightweight Multi-agent System for Automated Daily Safety Bulletin Generation in Construction” 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

Xianxiang Zhao, Advik Mehta, Falak Sethi, Brian Gue, Qipei Mei, Lingzi Wu

LM-DSB generates daily safety bulletins through a staged workflow that plans the briefing, retrieves safety guidance, and writes using a fine-tuned Llama 3-8B model. Across three scenarios, it scored higher than single-agent retrieval-augmented generation on all four text-overlap metrics, with a length ratio of 1.01 versus 0.39; this comparison does not isolate orchestration from fine-tuning or establish safety outcomes.

Summary by alphaXiv.

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