Reproduce “NEIGHBOR: Aligning Service Behaviour and Evidence Use in an LLM Agent for Community Resident Services” 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
Zongyou Yang, Yinghan Hou, Xiaokun Yang
NEIGHBOR trains service behaviours into a resident-service assistant and combines tool use with source facts, document retrieval, and read-only access to residents’ records. In an offline expert review of 50 held-out queries, it scored 7.74/10 versus 5.86 for the same-tools untuned agent, higher on 46 queries; the comparison does not isolate behaviour training from factual fine-tuning.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.