🛰️ Daily AI Frontier
‹ back to 2026-08-20

EVADE: Evidence-Verified Agentic Diagnosis with Escape

Research Medical/Healthcare AI

Ranking

Overall 77
Content 95
Popularity 35

Observed public metrics from 1 member.

Representative image for EVADE: Evidence-Verified Agentic Diagnosis with Escape

Merged summary

TL;DR - EVADE is a training-free diagnostic framework that makes a frozen medical vision-language model compare answers from full and self-localized zoomed image views, abstaining when they disagree. It improves calibration and selective safety without sacrificing accuracy.

  • On VQA-RAD, SLAKE, and PathVQA with Qwen2.5-VL-7B, EVADE reduced expected calibration error by up to 45% versus zero-shot.
  • Cross-view consistency avoids relying on textual self-verification, which can itself hallucinate.
  • EVADE was the only evaluated method to improve both calibration and selective risk while maintaining accuracy.
  • Self-proposed crops localized diagnostic structures better than center or random crops, but reliability gains came from agreement gating and abstention rather than answer revision.

Sources (1)

EVADE: Evidence-Verified Agentic Diagnosis with Escape

arXiv cs.CV Mohaimenul Azam Khan Raiaan, Nur Mohammad Fahad 2026-08-19 arXiv:2608.18833
Public signals Semantic Scholar citations 0 · Semantic Scholar influential citations 0
Providers: Hugging Face · N/A OpenAlex · N/A Publisher · N/A Semantic Scholar · Citations 0 · Influential citations 0 X · N/A Fetched 2026-08-22 14:14:59.326893 UTC

TL;DR - EVADE is a training-free diagnostic framework that makes a frozen medical vision-language model compare answers from full and self-localized zoomed image views, abstaining when they disagree. It improves calibration and selective safety without sacrificing accuracy.

  • On VQA-RAD, SLAKE, and PathVQA with Qwen2.5-VL-7B, EVADE reduced expected calibration error by up to 45% versus zero-shot.
  • Cross-view consistency avoids relying on textual self-verification, which can itself hallucinate.
  • EVADE was the only evaluated method to improve both calibration and selective risk while maintaining accuracy.
  • Self-proposed crops localized diagnostic structures better than center or random crops, but reliability gains came from agreement gating and abstention rather than answer revision.
item →