mlsanityData to 14 Oct 2026

Uncertainty Is Not a Safety Net for Clinical VQA, but Can It Anticipate Model Failure?

arXiv 2606.16583 ↗First posted 15 Jun 2026cs.CLcited by 0

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

Abstract (arXiv)

Safe deployment of clinical vision-language models (VLMs) requires reliable uncertainty estimation (UE): a signal indicating when predictions should be trusted or escalated to a clinician. We test whether current UE methods actually deliver this signal. Benchmarking 8 methods across 12 VLMs on clinical visual question-answering (VQA), we find that UE quality is not an intrinsic property of the UE method: it tracks model accuracy, degrading precisely where the model performance is weakest, and therefore where reliability is most needed. When we stress-test models by hiding the correct option among the multiple-choice answers (NOTA perturbations), accuracy collapses while uncertainty barely changes, leaving models systematically miscalibrated. Yet, we find that uncertainty on the unperturbed input reliably anticipates which predictions will collapse under NOTA, indicating that UE in current VLMs carries diagnostic information about model fragility. Our results position UE as a diagnostic tool for identifying fragile predictions and motivate perturbation-based evaluation as a path toward safe clinical deployment.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Medical vision-language models, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 17 of its 45 references that are in the map. First posted on arXiv 15 Jun 2026.

Builds on

17 of 45 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024Vision-Language and Large Language Model Performance in Gastroenterology: GPT, Claude, Llama, Phi, Mistral, Gemma, and Quantized Models25 Aug

Other references · 16

  1. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  2. 2022Large Language Models Encode Clinical Knowledgecited by 1,830
  3. 2019PubMedQA: A Dataset for Biomedical Research Question Answeringcited by 1,288
  4. 2024Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Modelscited by 329
  5. 2024INSIDE: LLMs' Internal States Retain the Power of Hallucination Detectioncited by 323
  6. 2025LLaVA-OneVision-1.5: Fully Open Framework for Democratized Multimodal Trainingcited by 244
  7. 2025MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understandingcited by 224
  8. 2021Medical Visual Question Answering: A Surveycited by 93
  9. 2024VL-Uncertainty: Detecting Hallucination in Large Vision-Language Model via Uncertainty Estimationcited by 73
  10. 2024MedExQA: Medical Question Answering Benchmark with Multiple Explanationscited by 60
  11. 2024Combining Confidence Elicitation and Sample-based Methods for Uncertainty Quantification in Misinformation Mitigationcited by 19
  12. 2024Certainly Uncertain: A Benchmark and Metric for Multimodal Epistemic and Aleatoric Awarenesscited by 14
4 more
  1. 2025ViLU: Learning Vision-Language Uncertainties for Failure Predictioncited by 9
  2. 2025Exploiting the Asymmetric Uncertainty Structure of Pre-trained VLMs on the Unit Hyperspherecited by 4
  3. 2026Calibrated? Not for Everyone: How Sexual Orientation and Religious Markers Distort LLM Accuracy and Confidence in Medical QAcited by 2
  4. 2025Unexplored flaws in multiple-choice VQA make benchmarking unreliablecited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024SURE-VQA: Systematic Understanding of Robustness Evaluation in Medical VQA Tasks
  2. 2026Rethinking Correctness for Uncertainty Estimation in Clinical Prediction with Vision-Language Models
  3. 2026Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation
  4. 2024Uncertainty-Aware Evaluation for Vision-Language Models
  5. 2025Improving VQA Reliability: A Dual-Assessment Approach with Self-Reflection and Cross-Model Verification
  6. 2025Mind the Gap: Benchmarking LLM Uncertainty and Calibration with Specialty-Aware Clinical QA and Reasoning-Based Behavioural Features
  7. 2025Uncertainty-Driven Expert Control: Enhancing the Reliability of Medical Vision-Language Models
  8. 2026Just how sure are you? Improving Verbalized Uncertainty Calibration in Medical VQA