Metraipolis › archive › dc_27eef27cbf45
CARE: Confidence-Aware Reasoning for Reliable Medical VQA — arXiv:2608.10964
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.10964
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-19T06:09:34.599Z
- SHA-256 of the archived text
16dda7e3f4b70a85b646c608f71c047441010daa49821fb1503bc976c7a2f8cd- Stable URL of this page
https://metraipolis.com/documento/dc_27eef27cbf45
How to cite
CARE: Confidence-Aware Reasoning for Reliable Medical VQA — arXiv:2608.10964 — arXiv (en), https://arxiv.org/abs/2608.10964 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_27eef27cbf45 (Metraipolis archive dc_27eef27cbf45, sha256 16dda7e3f4b70a85b646c608f71c047441010daa49821fb1503bc976c7a2f8cd)
BibTeX
@misc{dc27eef27cbf45,
title = {CARE: Confidence-Aware Reasoning for Reliable Medical VQA — arXiv:2608.10964},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.10964},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_27eef27cbf45; sha256 16dda7e3f4b70a85b646c608f71c047441010daa49821fb1503bc976c7a2f8cd},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.10964. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Reinforcement Fine-Tuning (RFT) has enabled medical Multimodal Large Language Models (MLLMs) to produce Chain-of-Thought (CoT) reasoning for visual question answering, yet these models suffer from $\textit{confidence miscalibration}$---a systematic gap between expressed certainty and actual diagnostic accuracy that undermines clinical trust. We propose $\textbf{CARE}$, a $\textbf{C}$onfidence-$\textbf{A}$ware medical $\textbf{RE}$asoning framework that jointly optimizes accuracy and calibration through a dual-stage pipeline. First, a scalable Medical-CoT synthesis provides structured cold-start data for Supervised Fine-Tuning. Second, Group Relative Policy Optimization (GRPO) with a novel $\textbf{Confidence-Aware Reward (CAR)}$ mechanism ties the model's confidence to diagnostic correctness within the reward signal. Across three Medical VQA benchmarks, $\textbf{CARE}$ achieves the highest diagnostic accuracy while obtaining the lowest Expected Calibration Error and Hallucination Rate, establishing a foundation for trustworthy clinical decision support. Our code is available at this https URL .
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-12 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.10964 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T05:57:28.418Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).