MARG: Multi-Agent Review Generation for Scientific Papers
This addresses the challenge of providing specific and helpful feedback for scientific papers, which is incremental as it builds on existing LLM capabilities.
The paper tackled the problem of generating feedback for scientific papers using LLMs, developing MARG, a multi-agent approach that reduces generic comments from 60% to 29% and increases good comments per paper from 1.7 to 3.7 (a 2.2x improvement).
We study the ability of LLMs to generate feedback for scientific papers and develop MARG, a feedback generation approach using multiple LLM instances that engage in internal discussion. By distributing paper text across agents, MARG can consume the full text of papers beyond the input length limitations of the base LLM, and by specializing agents and incorporating sub-tasks tailored to different comment types (experiments, clarity, impact) it improves the helpfulness and specificity of feedback. In a user study, baseline methods using GPT-4 were rated as producing generic or very generic comments more than half the time, and only 1.7 comments per paper were rated as good overall in the best baseline. Our system substantially improves the ability of GPT-4 to generate specific and helpful feedback, reducing the rate of generic comments from 60% to 29% and generating 3.7 good comments per paper (a 2.2x improvement).