Separating Capability from Confidence: Grounded Dual-State Calibration for GRPO-Trained Medical Vision-Language Models

2026-09-06

AI Summary

本文提出DualRead方法，通过分离能力学习与置信度估计，提高医学视觉-语言模型的正确性判别和校准，同时保持答案准确性。

Abstract

Medical vision-language models (VLMs) require confidence that reflects both answer correctness and patient-specific visual evidence. Recent GRPO-based methods optimize verbalized confidence together with answer generation. However, this joint optimization may interfere with answer learning and drive confidence toward near-binary values. Verbalized confidence also provides no explicit assessment of visual support. We therefore separate capability learning from confidence estimation and propose \textbf{DualRead}. DualRead builds on the insight that reliability can be read from the actor's internal states at critical moments in the answering process. It freezes the GRPO-trained actor and combines pre-answer solvability with a post-answer assessment of the generated answer and its visual support. To further assess whether confidence reflects visual grounding, we introduce \textbf{Counterfactual Confidence Grounding AUC} (CCG-AUC). It measures whether confidence decreases when real-image substitution changes the actor from correct to incorrect. Across two VLM backbones and both in- and out-of-distribution medical VQA benchmarks, DualRead improves correctness discrimination and calibration over verbalized confidence while preserving answer accuracy. CCG-AUC reveals whether confidence responds to answer-relevant visual evidence rather than primarily to non-visual cues.

Problem

• Medical Vision-Language Models
• Confidence Estimation
• Visual Evidence
• GRPO-Trained Models
• Calibration

Innovation

• DualRead
• GRPO
• Confidence Estimation
• Visual Support
• CCG-AUC