# MedQA-MM: Shortcuts Behind Medical Visual Reasoning

arXiv: https://arxiv.org/abs/2609.03261 · HTML: https://mlsanity.com/paper/2609.03261/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning.
- Builds on 6 of its 58 references that are in the map.
- First posted on arXiv 3 Sep 2026.

## Placed in

- LLM reasoning / [Clinical and biomedical applications](https://mlsanity.com/lr/clinical-and-biomedical-applications/)

## Builds on (most cited first, 6 of 6)

- [Qwen3-VL Technical Report](https://mlsanity.com/paper/2511.21631/) (2025)
- [MedGemma Technical Report](https://mlsanity.com/paper/2507.05201/) (2025)
- [Capabilities of Gemini Models in Medicine](https://mlsanity.com/paper/2404.18416/) (2024)
- [MIRAGE: The Illusion of Visual Understanding](https://mlsanity.com/paper/2603.21687/) (2026)
- [MedGemma 1.5 Technical Report](https://mlsanity.com/paper/2604.05081/) (2026)
- [MultiViewDx: Evidence-Linked Multi-View Clinical Diagnosis](https://mlsanity.com/paper/2410.14948/) (2024)

## Abstract (arXiv)

A benchmark score credits final answers, but not the route by which an item can be answered. In medical multimodal multiple-choice questions (MCQs), this distinction matters because a correct answer can be supported by the intended image finding or by benchmark-preserved cues in the wording of answers, non-visual clinical text, visible image text, artificial annotations, or device/context artifacts. We call the resulting score-level overinterpretation reasoning inflation. Here, a route is an observable input path that can support answer selection, not a claim about the model's hidden cognition. Across six medical multimodal MCQ datasets, we separate candidate cues from behavioral evidence through prompt- and image-side audits, modality ablations, and matched repairs that preserve the medical target and answer key. In a 13-configuration open-model panel, full-input accuracy is 62.63%, while text-only and options-only settings achieve 53.96% and 29.71%, respectively. Removing length-gap, absolute/conspicuous, and spatial/prepositional cues lowers accuracy by 6.58, 3.50, and 4.77 percentage points. We also construct MedQA-MM, a 1,000-item shortcut-mitigated subset, where text-only and options-only accuracy fall to 5.21% and 12.33%. This does not imply that models never use images; it shows that medical image-reasoning claims require route-level evidence.