Ensemble LLM Approach

Ensemble methods for large language models (LLMs) aim to improve performance and robustness by combining the predictions of multiple individual LLMs. Current research focuses on optimizing ensemble composition, often prioritizing diversity among component models to leverage their complementary strengths, and employing techniques like weighted averaging or learning-to-ensemble approaches to aggregate predictions effectively. This approach has demonstrated improved accuracy and efficiency across various NLP tasks, including question answering, text summarization, and even impacting real-world applications like e-commerce product attribute extraction and AI-generated text detection. The ongoing exploration of optimal ensemble size and the development of theoretically grounded ensemble algorithms are key areas of active investigation.

Papers

October 21, 2024

Large language models enabled multiagent ensemble method for efficient EHR data labeling
Jingwei Huang, Kuroush Nezafati, Ismael Villanueva-Miranda, Zifan Gu, Ann Marie Navar, Tingyi Wanyan, Qin Zhou, Bo Yao, Ruichen Rong, Xiaowei Zhan, Guanghua Xiao, Eric D. Peterson, Donghan M. Yang, Yang Xie
Large Language Model Ensemble LLM Approach Agent Ensemble

October 4, 2024

LLM-TOPLA: Efficient LLM Ensemble by Maximising Diversity
Selim Furkan Tekin, Fatih Ilhan, Tiansheng Huang, Sihao Hu, Ling Liu
Ensemble Learning Ensemble Diversity Ensemble LLM Approach Top $K$ Classification

February 29, 2024

LLM-Ensemble: Optimal Large Language Model Ensemble Method for E-commerce Product Attribute Value Extraction
Chenhao Fang, Xiaohan Li, Zezhong Fan, Jianpeng Xu, Kaushiki Nag, Evren Korpeoglu, Sushant Kumar, Kannan Achan
Language Model Attribute Value Extraction Product Attribute Value Extraction Ensemble LLM Approach

September 14, 2023

Generative AI Text Classification using Ensemble LLM Approaches
Harika Abburi, Michael Suesserman, Nirmala Pudota, Balaji Veeramani, Edward Bowen, Sanmitra Bhattacharya
Language Model Natural Language Processing Task Model Attribution Ensemble LLM Approach AI Generated Language

August 27, 2023

Leveraging Linear Independence of Component Classifiers: Optimizing Size and Prediction Accuracy for Online Ensembles
Enes Bektas, Fazli Can
Prediction Accuracy Size Matter Linear Dependence Sub Component Level Segmentation Effective Ensemble Ensemble LLM Approach

May 24, 2023

Are Pre-trained Language Models Useful for Model Ensemble in Chinese Grammatical Error Correction?
Chenming Tang, Xiuyu Wu, Yunfang Wu
Pre Trained Language Model Grammatical Error Correction Model Ensemble Chinese Grammatical Error Correction Ensemble LLM Approach

March 3, 2023

Evaluation of Confidence-based Ensembling in Deep Learning Image Classification
Rafael Rosales, Peter Popov, Michael Paulitsch
Global Evaluation Image Classification Diverse Ensemble ImageNet Dataset Model Ensembling Ensemble Machine Learning LLM Ensemble Ensemble LLM Approach

October 22, 2022

Diversity-Promoting Ensemble for Medical Image Segmentation
Mariana-Iuliana Georgescu, Radu Tudor Ionescu, Andreea-Iuliana Miron
Medical Image Segmentation Diverse Ensemble U Net Deep Learning System Gastrointestinal Tract Ensemble LLM Approach