Metraipolis › archive › dc_e78f79b02fa6
EVADE: Evidence-Verified Agentic Diagnosis with Escape — arXiv:2608.18833
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.18833
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T14:06:28.644Z
- SHA-256 of the archived text
8fe33a4fc4865fadcd4d4076d0cf3ca0a1f2db368f4c6a92bd439cf5fc197e7e- Stable URL of this page
https://metraipolis.com/documento/dc_e78f79b02fa6
How to cite
EVADE: Evidence-Verified Agentic Diagnosis with Escape — arXiv:2608.18833 — arXiv (en), https://arxiv.org/abs/2608.18833 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e78f79b02fa6 (Metraipolis archive dc_e78f79b02fa6, sha256 8fe33a4fc4865fadcd4d4076d0cf3ca0a1f2db368f4c6a92bd439cf5fc197e7e)
BibTeX
@misc{dce78f79b02fa6,
title = {EVADE: Evidence-Verified Agentic Diagnosis with Escape — arXiv:2608.18833},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.18833},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e78f79b02fa6; sha256 8fe33a4fc4865fadcd4d4076d0cf3ca0a1f2db368f4c6a92bd439cf5fc197e7e},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.18833. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Medical vision-language models (VLMs) can achieve high accuracy but remain unreliable: they are systematically overconfident, benefit little from test-time reasoning, and lack the ability to reliably calibrate trust in their own responses. We introduce EVADE (Evidence-Verified Agentic Diagnosis with Escape), an inferential, non-training method that enhances the safety of deploying a single frozen VLM. EVADE responds and, when uncertain, localises the region most diagnostically relevant, re-answers on a zoomed view, and commits only when both the entire image and the zoomed view responses agree; otherwise, it abstains. To directly address verification hallucination in single-model self-checking, our main idea is to verify gate consistency across different image views rather than re-reading the model's own text. Experimental evaluation on VQA-RAD, SLAKE, and PathVQA using Qwen2.5-VL-7B reports that EVADE is the only method that simultaneously improves both calibration and selective risk while maintaining accuracy, reducing expected calibration error (ECE) by up to 45% compared to zero-shot. Chain-of-thought, self-consistency, and self-verification all fail at least one axis. A grounding analysis reports that self-proposed regions perform better at diagnostic structure localisation than centres or random crops. However, a 7B VLM cannot use this localisation to revise answers. Therefore, reliability gains come from the consistency gate and calibrated abstention. Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-20 in cs.CV). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2608.18833 — 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 2026-10-18T14:00:22.582Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-20?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).