MARG: Multi-Agent Review Generation for Scientific Papers
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911751345274880 |
|---|---|
| author | D'Arcy, Mike Hope, Tom Birnbaum, Larry Downey, Doug |
| author_facet | D'Arcy, Mike Hope, Tom Birnbaum, Larry Downey, Doug |
| contents | 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). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_04259 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | MARG: Multi-Agent Review Generation for Scientific Papers D'Arcy, Mike Hope, Tom Birnbaum, Larry Downey, Doug Computation and Language 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). |
| title | MARG: Multi-Agent Review Generation for Scientific Papers |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2401.04259 |