Skip to content
AI Atlas

ModaLens: Measuring Image Sensitivity in Report-Conditioned Medical VLMs

Published 16 Sept 2026arXiv:2609.15635

data quality75

Updated 7 h ago · first seen 14 Sept 2026

paper_01M2JJGQ7RM7Q4TTY3J1RFA9PP

Abstract

A radiology report can already answer a clinical question, so it is hard to tell whether a vision-language model also uses the image. ModaLens, a paired image-swap audit, measures how report availability changes image sensitivity: MedGemma-27B on 3,199 paired MIMIC-CXR cases from 293 patients, all 14 questions per case (13 finding-specific and one composite), each image replaced by one from another study, usually of the same patient, with question and report fixed. Under an explicit answer instruction, the model's generated answer changes on 4.26 percent of trials with the report and 20.94 percent without it, a paired increase of 16.7 points (patient-clustered 95 percent CI 15.6 to 17.7), so report availability reduces image-swap sensitivity under this protocol; the original prompt with a lowercase first-token readout gives 4.70 percent against 17.07 percent, and substitutions also move continuous answer scores where the binary prediction does not change. The labels are derived from reports, which limits conclusions about visual correctness; the direction replicates in two further model lineages. Code, the exact prompts and a run record for every number are at https://github.com/criticaldata/MODALENS.

Authors

Authors 12

Anqi Peter LiAshley ZhangFelipe Ocampo OsorioKushul Reddy PalakalaLeo Anthony CeliMaximin LangeQuang BuiRafi Al AttrachSahil KapadiaSebasti\'an Andr\'es Cajas Ord\'o\~nezXinyue ZhangZakaria Laouabdia Sellami

Linked names open researcher pages (created from the paper's author list; name-only, no affiliation unless a source states it). Unlinked names have no researcher record yet.

Organizations

Organizations 0

No organization stated. arXiv metadata does not carry affiliations; an organization is linked only when a model card or lab page cites the paper.

Models

Models introduced or described 0

Inbound described_by relations from model cards and documentation.

No model links this paper yet

Model pages link papers through their model cards and documentation; the relation is written only when a source states it.

Datasets

Datasets used 0

No dataset relation recorded.

Benchmarks

Benchmarks used 0

No benchmark relation recorded.

Code

Repositories & frameworks 0

No repository linked.

Timeline

Timeline 6

Full timeline →

Sources

Sources 3

Source documents
SourceDocumentTypeTierLast observedSnapshots
arXiv (Atom API + RSS)rss.arxiv.org/rss/cs.CV feedT1· Official7 h ago4
arXiv (Atom API + RSS)rss.arxiv.org/rss/cs.AI feedT1· Official7 h ago5
Hugging Face Hub (public pages, model cards, papers)huggingface.co/papers listingT2· Quality secondary4 h ago30

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.