Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation

• URL: http://arxiv.org/abs/2604.02543v1
• Date: Thu, 02 Apr 2026 21:52:29 GMT
• Title: Overconfidence and Calibration in Medical VQA: Empirical Findings and Hallucination-Aware Mitigation
• Abstract summary: Overconfidence in vision-language models (VLMs) persists across model families and is not resolved by scaling or prompting.<n>Post-hoc calibration approaches, such as Platt scaling, reduce calibration error and consistently outperform the prompt-based strategy.<n> hallucination-aware calibration (HAC) incorporates vision-grounded hallucination detection signals as complementary inputs to refine confidence estimates.
• Score: 4.247706336083815
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: As vision-language models (VLMs) are increasingly deployed in clinical decision support, more than accuracy is required: knowing when to trust their predictions is equally critical. Yet, a comprehensive and systematic investigation into the overconfidence of these models remains notably scarce in the medical domain. We address this gap through a comprehensive empirical study of confidence calibration in VLMs, spanning three model families (Qwen3-VL, InternVL3, LLaVA-NeXT), three model scales (2B--38B), and multiple confidence estimation prompting strategies, across three medical visual question answering (VQA) benchmarks. Our study yields three key findings: First, overconfidence persists across model families and is not resolved by scaling or prompting, such as chain-of-thought and verbalized confidence variants. Second, simple post-hoc calibration approaches, such as Platt scaling, reduce calibration error and consistently outperform the prompt-based strategy. Third, due to their (strict) monotonicity, these post-hoc calibration methods are inherently limited in improving the discriminative quality of predictions, leaving AUROC at the same level. Motivated by these findings, we investigate hallucination-aware calibration (HAC), which incorporates vision-grounded hallucination detection signals as complementary inputs to refine confidence estimates. We find that leveraging these hallucination signals improves both calibration and AUROC, with the largest gains on open-ended questions. Overall, our findings suggest post-hoc calibration as standard practice for medical VLM deployment over raw confidence estimates, and highlight the practical usefulness of hallucination signals to enable more reliable use of VLMs in medical VQA.

Related papers

• When Confidence Fails: Overconfidence in LLMs under Uncertainty and Missing Clinical Information [14.659028787864996]
  Large Language Models (LLMs) have achieved strong performance in medical question answering and clinical reasoning tasks.<n>We conduct a behavioral analysis of LLMs under clinical information uncertainty.<n>We observe significant variation across models in their ability to abstain when the correct answer is unavailable.
  arXiv  Detail & Related papers  (2026-08-10T03:28:46Z)
• Discovery of Hidden Miscalibration Regimes [52.452902154360565]
  Models may be systematically overconfident on some kinds of inputs and underconfident on others.<n>We define the corresponding miscalibration field and propose a diagnostic framework for estimating it.<n>Our approach learns a calibration-aware representation of the input space and estimates signed local miscalibration by kernel smoothing in the learned geometry.
  arXiv  Detail & Related papers  (2026-05-13T13:07:50Z)
• Benchmarking Multi-turn Medical Diagnosis: Hold, Lure, and Self-Correction [72.89352076103889]
  Large language models (LLMs) achieve high accuracy in medical diagnosis when all clinical information is provided in a single turn.<n>We introduce MINT, a high-fidelity, multi-turn medical diagnosis benchmark comprising 1,035 cases.<n>We uncover three persistent behavioral patterns that significantly impact diagnostic decisions.
  arXiv  Detail & Related papers  (2026-04-06T00:23:10Z)
• Guideline-Grounded Evidence Accumulation for High-Stakes Agent Verification [60.18369393468405]
  Existing verifiers usually underperform owing to a lack of domain knowledge and limited calibration.<n>GLEAN compiles expert-curated protocols into trajectory-informed, well-calibrated correctness signals.<n>We empirically validate GLEAN with agentic clinical diagnosis across three diseases from the MIMIC-IV dataset.
  arXiv  Detail & Related papers  (2026-03-03T09:36:43Z)
• On Calibration of Large Language Models: From Response To Capability [66.59139960234326]
  Large language models (LLMs) are widely deployed as general-purpose problem solvers.<n>We introduce capability calibration, which targets the model's expected accuracy on a query.<n>Our results demonstrate that capability-calibrated confidence improves pass@$k$ prediction and inference budget allocation.
  arXiv  Detail & Related papers  (2026-02-14T01:07:45Z)
• Towards Reliable Medical LLMs: Benchmarking and Enhancing Confidence Estimation of Large Language Models in Medical Consultation [97.36081721024728]
  We propose the first benchmark for assessing confidence in multi-turn interaction during realistic medical consultations.<n>Our benchmark unifies three types of medical data for open-ended diagnostic generation.<n>We present MedConf, an evidence-grounded linguistic self-assessment framework.
  arXiv  Detail & Related papers  (2026-01-22T04:51:39Z)
• Benchmarking Egocentric Clinical Intent Understanding Capability for Medical Multimodal Large Language Models [48.95516224614331]
  We introduce MedGaze-Bench, the first benchmark leveraging clinician gaze as a Cognitive Cursor to assess intent understanding across surgery, emergency simulation, and diagnostic interpretation.<n>Our benchmark addresses three fundamental challenges: visual homogeneity of anatomical structures, strict temporal-causal dependencies in clinical, and implicit adherence to safety protocols.
  arXiv  Detail & Related papers  (2026-01-11T02:20:40Z)
• Refine and Align: Confidence Calibration through Multi-Agent Interaction in VQA [8.879640496857661]
  We introduce AlignVQA, a debate-based multi-agent framework for Visual Question Answering.<n>We find that more calibrated specialized agents produce better aligned confidences.<n>We also introduce a novel differentiable calibration-aware loss function called aligncal to fine-tune the specialized agents.
  arXiv  Detail & Related papers  (2025-11-14T11:08:21Z)
• A Multi-faceted Analysis of Cognitive Abilities: Evaluating Prompt Methods with Large Language Models on the CONSORT Checklist [1.1731001328350983]
  This study applies a behavioral and metacognitive analytic approach using an expert-validated dataset.<n>We analyze both cognitive adaptation and calibration error using metrics: Expected Error (ECE) and a baseline-normalized Relative Error (RCE)<n>Our results reveal pronounced miscalibration and overconfidence in both models, especially under clinical role-playing conditions.
  arXiv  Detail & Related papers  (2025-10-22T00:15:02Z)
• Calibration-Aware Prompt Learning for Medical Vision-Language Models [44.97741487992985]
  Miscalibrated predictions can lead to overconfident errors, undermining clinical trust and decision-making reliability.<n>We introduce CalibPrompt, the first framework to calibrate Med-VLMs during prompt tuning.<n>CalibPrompt consistently improves calibration without drastically affecting clean accuracy.
  arXiv  Detail & Related papers  (2025-09-18T17:59:58Z)
• Uncertainty-Driven Expert Control: Enhancing the Reliability of Medical Vision-Language Models [52.2001050216955]
  Existing methods aim to enhance the performance of Medical Vision Language Model (MedVLM) by adjusting model structure, fine-tuning with high-quality data, or through preference fine-tuning.<n>We propose an expert-in-the-loop framework named Expert-Controlled-Free Guidance (Expert-CFG) to align MedVLM with clinical expertise without additional training.
  arXiv  Detail & Related papers  (2025-07-12T09:03:30Z)
• Fact-Level Confidence Calibration and Self-Correction [64.40105513819272]
  We propose a Fact-Level framework that calibrates confidence to relevance-weighted correctness at the fact level. We also develop Confidence-Guided Fact-level Self-Correction ($textbfConFix$), which uses high-confidence facts within a response as additional knowledge to improve low-confidence ones.
  arXiv  Detail & Related papers  (2024-11-20T14:15:18Z)
• Enhancing Healthcare LLM Trust with Atypical Presentations Recalibration [20.049443396032423]
  Black-box large language models (LLMs) are increasingly deployed in various environments. LLMs often exhibit overconfidence, leading to potential risks and misjudgments. We propose a novel method, textitAtypical presentations Recalibration, which leverages atypical presentations to adjust the model's confidence estimates.
  arXiv  Detail & Related papers  (2024-09-05T03:45:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.