LLM-Collaboration on Automatic Science Journalism for the General Audience

TL;DR

This paper introduces a collaborative framework of three open-source LLMs mimicking the scientific journalism process, resulting in more accessible science articles for the general public compared to existing methods.

Contribution

It presents a novel multi-LLM collaboration framework that simulates the writing, reading, feedback, and revision workflow for science journalism.

Findings

Generated articles are more accessible than existing methods.

Collaboration of small open-source LLMs can outperform GPT-4 in accessibility.

The framework effectively mimics real-world science journalism processes.

Abstract

Science journalism reports current scientific discoveries to non-specialists, aiming to enable public comprehension of the state of the art. However, this task can be challenging as the audience often lacks specific knowledge about the presented research. To address this challenge, we propose a framework that integrates three LLMs mimicking the real-world writing-reading-feedback-revision workflow, with one LLM acting as the journalist, a smaller LLM as the general public reader, and the third LLM as an editor. The journalist's writing is iteratively refined by feedback from the reader and suggestions from the editor. Our experiments demonstrate that by leveraging the collaboration of two 7B and one 1.8B open-source LLMs, we can generate articles that are more accessible than those generated by existing methods, including advanced models such as GPT-4.