# Coling-UniA at SciVQA 2025: Few-Shot Example Retrieval and Confidence-Informed Ensembling for Multimodal Large Language Models

Source: [Coling-UniA at SciVQA 2025: Few-Shot Example Retrieval and Confidence-Informed Ensembling for Multimodal Large Language Models](<https://www.sota2.com/research/paper/68fa79c81e209131cea7f594>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T05:05:34.469Z

Authors: Christian Jaumann, Annemarie Friedrich, Rainer Lienhart

Published: 2025-07-03T06:43:05.000Z

## Abstract

This paper describes our system for the SciVQA 2025 Shared Task on Scientific Visual Question Answering. Our system employs an ensemble of two Multimodal Large Language Models and various few-shot example retrieval strategies. The model and few-shot setting are selected based on the figure and question type. We also select answers based on the models' confidence levels. On the blind test data, our system ranks third out of seven with an average F1 score of 85.12 across ROUGE-1, ROUGE-L, and BERTS. Our code is publicly available.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2507.02357>)

[PDF](<https://arxiv.org/pdf/2507.02357.pdf>)
