Multi-LLM Text Summarization
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916670703927296 |
|---|---|
| author | Fang, Jiangnan Liu, Cheng-Tse Kim, Jieun Bhedaru, Yash Liu, Ethan Singh, Nikhil Lipka, Nedim Mathur, Puneet Ahmed, Nesreen K. Dernoncourt, Franck Rossi, Ryan A. Deilamsalehy, Hanieh |
| author_facet | Fang, Jiangnan Liu, Cheng-Tse Kim, Jieun Bhedaru, Yash Liu, Ethan Singh, Nikhil Lipka, Nedim Mathur, Puneet Ahmed, Nesreen K. Dernoncourt, Franck Rossi, Ryan A. Deilamsalehy, Hanieh |
| contents | In this work, we propose a Multi-LLM summarization framework, and investigate two different multi-LLM strategies including centralized and decentralized. Our multi-LLM summarization framework has two fundamentally important steps at each round of conversation: generation and evaluation. These steps are different depending on whether our multi-LLM decentralized summarization is used or centralized. In both our multi-LLM decentralized and centralized strategies, we have k different LLMs that generate diverse summaries of the text. However, during evaluation, our multi-LLM centralized summarization approach leverages a single LLM to evaluate the summaries and select the best one whereas k LLMs are used for decentralized multi-LLM summarization. Overall, we find that our multi-LLM summarization approaches significantly outperform the baselines that leverage only a single LLM by up to 3x. These results indicate the effectiveness of multi-LLM approaches for summarization. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_15487 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Multi-LLM Text Summarization Fang, Jiangnan Liu, Cheng-Tse Kim, Jieun Bhedaru, Yash Liu, Ethan Singh, Nikhil Lipka, Nedim Mathur, Puneet Ahmed, Nesreen K. Dernoncourt, Franck Rossi, Ryan A. Deilamsalehy, Hanieh Computation and Language In this work, we propose a Multi-LLM summarization framework, and investigate two different multi-LLM strategies including centralized and decentralized. Our multi-LLM summarization framework has two fundamentally important steps at each round of conversation: generation and evaluation. These steps are different depending on whether our multi-LLM decentralized summarization is used or centralized. In both our multi-LLM decentralized and centralized strategies, we have k different LLMs that generate diverse summaries of the text. However, during evaluation, our multi-LLM centralized summarization approach leverages a single LLM to evaluate the summaries and select the best one whereas k LLMs are used for decentralized multi-LLM summarization. Overall, we find that our multi-LLM summarization approaches significantly outperform the baselines that leverage only a single LLM by up to 3x. These results indicate the effectiveness of multi-LLM approaches for summarization. |
| title | Multi-LLM Text Summarization |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2412.15487 |