Fusing Models with Complementary Expertise
It addresses the challenge of handling heterogeneous data at test time for AI models, which is an incremental improvement over existing expert model approaches.
The paper tackles the Fusion of Experts (FoE) problem by fusing outputs of expert models with complementary knowledge to improve generalization across tasks and domains, resulting in significant performance gains in image and text classification, text summarization, multiple-choice QA, and automatic text evaluation.
Training AI models that generalize across tasks and domains has long been among the open problems driving AI research. The emergence of Foundation Models made it easier to obtain expert models for a given task, but the heterogeneity of data that may be encountered at test time often means that any single expert is insufficient. We consider the Fusion of Experts (FoE) problem of fusing outputs of expert models with complementary knowledge of the data distribution and formulate it as an instance of supervised learning. Our method is applicable to both discriminative and generative tasks and leads to significant performance improvements in image and text classification, text summarization, multiple-choice QA, and automatic evaluation of generated text. We also extend our method to the "frugal" setting where it is desired to reduce the number of expert model evaluations at test time. Our implementation is publicly available at https://github.com/hwang595/FoE-ICLR2024.