What AI says about TreeReview

TreeReview is an LLM-based framework for deep and efficient scientific peer review that generates comprehensive reviews through a dynamic tree of questions and a hierarchical question-answering process. It provides implementation code and a benchmark dataset, with setup instructions (Conda environment, Python 3.11) and a command-line workflow to generate a review from a paper ID and an MMD file. The project aims to enable structured, scalable peer review by guiding large language models through organized questions to assess scientific papers.