Reproduce “Contri(e)ve: Context + Retrieve for Scholarly Question Answering” 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

Kanchan Shivashankar, Nadine Steinmetz

The authors build a scholarly question-answering system combining DBLP and SemOpenAlex knowledge graphs with Wikipedia text. Using a DBLP author ID, it gathers author-linked records and text, then supplies this context to Llama 3.1 8B-Instruct. On 702 Scholarly-QALD questions, it scored F1 0.40702 and exact-match accuracy 32.05%, but answers were inconsistent even when evidence was supplied.

Summary by alphaXiv.

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