LinguaMark: Do Multimodal Models Speak Fairly? A Benchmark-Based Evaluation
This work addresses fairness issues in LMMs for multilingual applications, but it is incremental as it builds on existing multimodal evaluation by focusing on multilingual capabilities.
The authors tackled the problem of biased and unfair outputs in Large Multimodal Models (LMMs) across languages by introducing LinguaMark, a benchmark for multilingual Visual Question Answering evaluation, finding that closed-source models like GPT-4o and Gemini2.5 achieved the highest overall performance, with Qwen2.5 showing strong generalization across multiple languages.
Large Multimodal Models (LMMs) are typically trained on vast corpora of image-text data but are often limited in linguistic coverage, leading to biased and unfair outputs across languages. While prior work has explored multimodal evaluation, less emphasis has been placed on assessing multilingual capabilities. In this work, we introduce LinguaMark, a benchmark designed to evaluate state-of-the-art LMMs on a multilingual Visual Question Answering (VQA) task. Our dataset comprises 6,875 image-text pairs spanning 11 languages and five social attributes. We evaluate models using three key metrics: Bias, Answer Relevancy, and Faithfulness. Our findings reveal that closed-source models generally achieve the highest overall performance. Both closed-source (GPT-4o and Gemini2.5) and open-source models (Gemma3, Qwen2.5) perform competitively across social attributes, and Qwen2.5 demonstrates strong generalization across multiple languages. We release our benchmark and evaluation code to encourage reproducibility and further research.