Reproduce “AiReview: An Open Platform for Accelerating Systematic Reviews with LLMs” 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

Xinyu Mao, Teerapong Leelanupab, Martin Potthast, Harrisen Scells, Guido Zuccon

An open-source platform, AiReview, integrates Large Language Models (LLMs) into the title and abstract screening phase of systematic reviews, providing users with transparent control over LLM parameters and prompts. This aims to accelerate evidence synthesis by offering flexible LLM assistance across various review stages.

Key insights

  • LLMs can be systematically integrated into systematic review screening at different stages of the workflow, offering varied levels of assistance and interaction.
  • Providing explicit, transparent control over LLM behavior, especially through editable prompts, is crucial for user trust, reproducibility, and mitigating the 'black box' problem in AI-assisted research.
  • A flexible framework for LLM integration can cater to diverse systematic review needs and team compositions, allowing users to balance efficiency gains with the potential for LLM-introduced bias.

Summary by alphaXiv.

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