mlsanityData to 7 Oct 2026

Confident but Unreliable: A Behavioral Safety Audit of Vision-Language Models on Brain MRI

arXiv 2608.02790 ↗First posted 3 Aug 2026cs.CVcited by 0

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

Abstract (arXiv)

Vision-language models (VLMs), including medical specialists, are increasingly proposed for medical imaging, yet their stated confidence is rarely evaluated separately from correctness. We use brain MRI as a controlled, high-stakes testbed for a broader failure mode in frontier multimodal systems: models can appear competent while lacking reliable self-knowledge. We present an automatically graded behavioral audit and pilot study of six instruction-tuned VLMs (five general-purpose and one medical specialist) on 4,102 images (4,032 axial/coronal/sagittal MRI slices from 250 subjects plus 70 non-brain/noise controls), with labels derived from public metadata and released expert segmentation masks rather than new human annotation. Across models, answer coverage is near-complete, but verbalized-confidence calibration is poor: ECE ranges from 0.27 to 0.40, mean confidence on incorrect answers ranges from 0.82 to 0.97, and 33-46% of answered items are high-confidence errors. The most accurate model is also the most confident on its errors, while a base/specialist family contrast suggests that medical adaptation improves tumor-presence detection without improving confidence reliability. Open-ended diagnostics further show that hallucination and abstention vary separately from multiple-choice accuracy. These findings argue that medical-image VLM evaluation should report verbalized-confidence reliability, confident error, hallucination, and abstention alongside accuracy.

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 7 of its 16 references that are in the map. First posted on arXiv 3 Aug 2026.

Builds on

7 of 16 references.

Closest ancestors · 3

Same direction, newest first

  1. 2026Calibrated Triage, Not Autonomy: Confidence Estimation for Medical Vision-Language Models14 Jun
  2. 2026Lost in the Hype: Revealing and Dissecting the Performance Degradation of Medical Multimodal Large Language Models in Image Classification9 Apr
  3. 2026Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation2 Apr

Other references · 4

  1. 2017On Calibration of Modern Neural Networkscited by 4,860
  2. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,696
  3. 2024Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scalingcited by 1,573
  4. 2024The 2024 Brain Tumor Segmentation (BraTS) Challenge: Glioma Segmentation on Post-treatment MRIcited by 65

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Prompts Without Evidence: How Neuroimaging Mentions Shift Clinical Vision-Language Model Predictions
  2. 2026Evaluating the Diagnostic Robustness of Vision-Language Models Under Visual and Textual Perturbations
  3. 2026Separating Capability from Confidence: Grounded Dual-State Calibration for GRPO-Trained Medical Vision-Language Models
  4. 2025On the Robustness of Medical Vision-Language Models: Are they Truly Generalizable?
  5. 2025EchoBench: Benchmarking Sycophancy in Medical Large Vision-Language Models
  6. 2026Toward Guarantees for Clinical Reasoning in Vision Language Models via Formal Verification
  7. 2026Consistent but Dangerous: Per-Sample Safety Classification Reveals False Reliability in Medical Vision-Language Models
  8. 2026NeuroVLM-Bench: Evaluation of Vision-Enabled Large Language Models for Clinical Reasoning in Neurological Disorders