Chart-HQA: A Benchmark for Hypothetical Question Answering in Charts
This addresses the issue of output biases in MLLMs for chart-based question answering, providing a new benchmark to improve model reasoning, though it is incremental as it builds on existing chart benchmarks.
The authors tackled the problem of multimodal large language models (MLLMs) relying on parametric memory rather than genuine chart understanding by introducing a Chart Hypothetical Question Answering (HQA) task and a benchmark called Chart-HQA, which revealed significant generalization challenges and imbalanced reasoning performance in 18 evaluated MLLMs.
Multimodal Large Language Models (MLLMs) have garnered significant attention for their strong visual-semantic understanding. Most existing chart benchmarks evaluate MLLMs' ability to parse information from charts to answer questions. However, they overlook the inherent output biases of MLLMs, where models rely on their parametric memory to answer questions rather than genuinely understanding the chart content. To address this limitation, we introduce a novel Chart Hypothetical Question Answering (HQA) task, which imposes assumptions on the same question to compel models to engage in counterfactual reasoning based on the chart content. Furthermore, we introduce HAI, a human-AI interactive data synthesis approach that leverages the efficient text-editing capabilities of LLMs alongside human expert knowledge to generate diverse and high-quality HQA data at a low cost. Using HAI, we construct Chart-HQA, a challenging benchmark synthesized from publicly available data sources. Evaluation results on 18 MLLMs of varying model sizes reveal that current models face significant generalization challenges and exhibit imbalanced reasoning performance on the HQA task.