Detecting Clinical Hallucinations in LVLMs via Counterfactual Visual Grounding Uncertainty
On the map: Multimodal LLMs / vision-language models → Hallucination mitigation in LVLMs
Abstract (arXiv)
Large vision-language models (LVLMs) are increasingly used for clinical image understanding, yet they remain vulnerable to hallucinations--producing textual findings or attributes not supported by the image. We present a vision-traceable hallucination detection framework that audits arbitrary LVLM responses via visual evidence grounding, requiring neither modification nor internal access to the hidden states of LVLMs. Given an LVLM response, we extract visually verifiable entities and use a medical-domain-adapted Qwen-VL grounding verifier to localize each entity on the input image. To enhance the robustness of our detection method, we introduce a counterfactual entity perturbation method and estimate visual evidence uncertainty by contrasting factual and counterfactual grounding results. Specifically, we compute an entity-level uncertainty score from the positive confidence, counterfactual confidence, and their grounding overlap for binary hallucination decision-making. Experiments on multiple medical imaging modalities and LVLM backbones demonstrate that our method consistently improves hallucination detection performance over recent baselines, while providing interpretable localization evidence and strong cross-model transferability. Code and dataset are available at https://github.com/Agentic-CliniAI/CounterVHD.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Hallucination mitigation in LVLMs, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 13 of its 15 references that are in the map. First posted on arXiv 26 Jun 2026.
Builds on
13 of 15 references.
Closest ancestors · 7
Same direction, newest first
- 2025Hallucination-Aware Multimodal Benchmark for Gastrointestinal Image Analysis with Large Vision-Language Models11 May
- 2025UniVRSE: Unified Vision-conditioned Response Semantic Entropy for Hallucination Detection in Medical Vision-Language Models26 Mar
- 2024Detecting and Evaluating Medical Hallucinations in Large Vision Language Models14 Jun
- 2024Detecting and Mitigating Hallucination in Large Vision Language Models via Fine-Grained AI Feedback22 Apr
- 2024Unified Hallucination Detection for Multimodal Large Language Models5 Feb
- 2023AMBER: An LLM-free Multi-dimensional Benchmark for MLLMs Hallucination Evaluation13 Nov
- 2023FaithScore: Fine-grained Evaluations of Hallucinations in Large Vision-Language Models2 Nov
Other references · 6
- 2025Qwen3-VL Technical Reportcited by 2,793
- 2025MedGemma Technical Reportcited by 436
- 2024Interactive Medical Image Segmentation: A Benchmark Dataset and Baselinecited by 30
- 2024Uncertainty-aware Medical Diagnostic Phrase Identification and Groundingcited by 13
- 2024ReXTrust: A Model for Fine-Grained Hallucination Detection in AI-Generated Radiology Reportscited by 12
- 2023Rethinking Radiology Report Generation via Causal Inspired Counterfactual Augmentationcited by 3
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026Deterministic Hallucination Detection in Medical VQA via Confidence-Evidence Bayesian Gain
- 2024Reference-free Hallucination Detection for Large Vision-Language Models
- 2026Hallucination Detection and Correction in Medical VLMs via Counter-Evidence Verification
- 2026VGS-Decoding: Visual Grounding Score Guided Decoding for Hallucination Mitigation in Medical VLMs
- 2024VL-Uncertainty: Detecting Hallucination in Large Vision-Language Model via Uncertainty Estimation
- 2024MedVH: Towards Systematic Evaluation of Hallucination for Large Vision Language Models in the Medical Context
- 2026UHP Detection: LVLMs have their Unique Hallucination Pattern in the Consistency Space
- 2026Beyond the Global Scores: Fine-Grained Token Grounding as a Robust Detector of LVLM Hallucinations