Black-box Model Ensembling for Textual and Visual Question Answering via Information Fusion
This addresses the problem of cost and accessibility in using black-box models for textual and visual question answering, though it is incremental as it builds on existing ensemble techniques.
The paper tackles the challenge of fine-tuning black-box models like LLMs and VQA models by introducing InfoSel, an ensemble method that dynamically selects the best model for predictions, achieving up to a +5.19% F1-score improvement with only 1K training instances.
A diverse range of large language models (LLMs), e.g., ChatGPT, and visual question answering (VQA) models, e.g., BLIP, have been developed for solving textual and visual question answering tasks. However, fine-tuning these models is either difficult, as it requires access via APIs, rendering them as black-boxes, or costly due to the need of tuning a large number of parameters. To address this, we introduce InfoSel, a data-efficient ensemble method that learns to dynamically pick the winner from existing black-box models for predictions on both textual and multimodal visual question answering tasks. Unlike traditional ensemble models, InfoSel does not rely on prediction probabilities or confidences, which typically are not available in black-box models. Experimental results on four datasets demonstrate that our approach achieves an absolute increase of up to +5.19\% in the F1-score compared to standalone LLMs using only 1K training instances.