Protecting multimodal large language models against misleading visualizations

Abstract: Visualizations play a pivotal role in daily communication in an increasingly datadriven world. Research on multimodal LLMs (MLLMs) for automated chart understanding has accelerated massively, with steady improvements on standard benchmarks. However, for MLLMs to be reliable, they must be robust to misleading visualizations, i.e., charts that distort the underlying data, leading readers to draw inaccurate conclusions that may support disinformation. Here, we uncover an important vulnerability: MLLM questionanswering (QA) accuracy on misleading visualizations drops on average to the level of the random baseline. To address this, we introduce the first inference-time methods to improve QA performance on misleading visualizations, without compromising accuracy on non-misleading ones. We find that two methods, table-based QA and redrawing the visualization, are effective, with improvements of up to 19.6 percentage points. We make our code and data available.