Redundancy Principles for MLLMs Benchmarks

TL;DR

This paper critically examines the redundancy in MLLM benchmarks across multiple dimensions, providing insights and strategies to improve benchmark diversity and effectiveness amid rapid growth.

Contribution

It introduces a comprehensive analysis of redundancy in MLLM benchmarks and proposes principles to guide the development of more effective and less redundant benchmarks.

Findings

Quantitative measurement of redundancy levels in existing benchmarks

Identification of key sources of redundancy in MLLM evaluations

Guidelines for constructing more diverse and effective benchmarks

Abstract

With the rapid iteration of Multi-modality Large Language Models (MLLMs) and the evolving demands of the field, the number of benchmarks produced annually has surged into the hundreds. The rapid growth has inevitably led to significant redundancy among benchmarks. Therefore, it is crucial to take a step back and critically assess the current state of redundancy and propose targeted principles for constructing effective MLLM benchmarks. In this paper, we focus on redundancy from three key perspectives: 1) Redundancy of benchmark capability dimensions, 2) Redundancy in the number of test questions, and 3) Cross-benchmark redundancy within specific domains. Through the comprehensive analysis over hundreds of MLLMs' performance across more than 20 benchmarks, we aim to quantitatively measure the level of redundancy lies in existing MLLM evaluations, provide valuable insights to guide the future development of MLLM benchmarks, and offer strategies to refine and address redundancy issues effectively. The code is available at https://github.com/zzc-1998/Benchmark-Redundancy.