mlsanityData to 7 Oct 2026

Polish Medical Visual Question Answering: Vision-Language Models Underutilize Visual Evidence

arXiv 2608.12928 ↗First posted 13 Aug 2026cs.AIcited by 0

On the map: Multimodal LLMs / vision-language models → Medical vision-language models

Abstract (arXiv)

We introduce a Polish-language medical visual question answering (VQA) benchmark, built from Polish Board Certification Examination questions for licensed physicians and dentists pursuing specialist certification. The benchmark comprises image-containing questions spanning diverse medical specialties and visual domains, together with a text-only question answering (QA) control set. We evaluate Polish-oriented, general-purpose open-weight, and commercial vision-language models. The task remains challenging: the best model achieves 79.0% accuracy on the full VQA set, and only GPT-5.6 surpasses the approximate human reference on the subset with available candidate responses; all other evaluated models perform worse than humans. To assess visual grounding, we compare complete inputs with configurations omitting the image, the question, or both, and categorize questions by image importance. Models derive more useful information from the question text than from the image and perform worse on image-dominant questions. Across both QA and VQA, they nevertheless achieve above-chance accuracy from the answer choices alone, showing that non-trivial performance can persist even when key task components are missing.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Medical vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 17 of its 37 references that are in the map. First posted on arXiv 13 Aug 2026.

Builds on

17 of 37 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026Does Language Shift Break Medical Vision-Language Models? Indonesian Radiology Visual Question Answering Case Study2 Jun

Other references · 16

  1. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,498
  2. 2026DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligencecited by 877
  3. 2021SLAKE: A Semantically-Labeled Knowledge-Enhanced Dataset for Medical Visual Question Answeringcited by 444
  4. 2017Don't Just Assume; Look and Answer: Overcoming Priors for Visual Question Answeringcited by 337
  5. 2019RUBi: Reducing Unimodal Biases in Visual Question Answeringcited by 183
  6. 2018Overcoming Language Priors in Visual Question Answering with Adversarial Regularizationcited by 104
  7. 2024Artifacts or Abduction: How Do LLMs Answer Multiple-Choice Questions Without the Question?cited by 68
  8. 2025Which of These Best Describes Multiple Choice Evaluation with LLMs? A) Forced B) Flawed C) Fixable D) All of the Abovecited by 51
  9. 2020Reducing Language Biases in Visual Question Answering with Visually-Grounded Question Encodercited by 34
  10. 2021Towards Developing a Multilingual and Code-Mixed Visual Question Answering System by Knowledge Distillationcited by 13
  11. 2024GPT-4 passes most of the 297 written Polish Board Certification Examinationscited by 10
  12. 2024Polish-English medical knowledge transfer: A new benchmark and resultscited by 8
4 more
  1. 2025LLMzSzL: a comprehensive LLM benchmark for Polishcited by 7
  2. 2026Annotation-Efficient Vision-Language Model Adaptation to the Polish Language Using the LLaVA Frameworkcited by 3
  3. 2026Reassessing High-Performing LLMs on Polish Medical Exams: True Competence or Bias-Driven Performance?cited by 1
  4. 2026Jako Tako or Fluent? Presenting PoVisLE: A Polish Vision-Language Evaluationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Are Large Vision Language Models Truly Grounded in Medical Images? Evidence from Italian Clinical Visual Question Answering
  2. 2024Fusion of Domain-Adapted Vision and Language Models for Medical Visual Question Answering
  3. 2023Visual Question Answering in the Medical Domain
  4. 2024Targeted Visual Prompting for Medical Visual Question Answering
  5. 2024OmniMedVQA: A New Large-Scale Comprehensive Evaluation Benchmark for Medical LVLM
  6. 2026Do Medical Vision Language Models Actually See? A Counterfactual Grounding Framework and Hard-Negative Contrastive Training for Visually-Reliant Medical VLMs
  7. 2025CoTBox-TTT: Grounding Medical VQA with Visual Chain-of-Thought Boxes During Test-time Training
  8. 2025Adapting Lightweight Vision Language Models for Radiological Visual Question Answering