Diversity-Aware Ensembling of Language Models Based on Topological Data Analysis

arXiv:2402.14184v1 Announce Type: new
Abstract: Ensembles are important tools for improving the performance of machine learning models. In cases related to natural language processing, ensembles boost the performance of a method due to multiple large models available in open source. However, existing approaches mostly rely on simple averaging of predictions by ensembles with equal weights for each model, ignoring differences in the quality and conformity of models. We propose to estimate weights for ensembles of NLP models using not only …

abstract analysis arxiv boost cases cs.cl cs.lg data data analysis diversity language language models language processing large models machine machine learning machine learning models multiple natural natural language natural language processing open source performance predictions processing simple tools type