AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

ICYM2I: The illusion of multimodal informativeness under missingness

Young Sang Choi, Vincent Jeanselme, Pierre Elias, Shalmali Joshi

ICLR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Multimodal fusion with missing modalities · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
apply IPW correction to training and evaluationnaive complete-case analysis · Bitwise logic AND (MAR)AUROC of X1 model: 0.66 -> 0.83Improved
apply IPW correctionnaive complete-case analysis · Bitwise logic OR (MAR)AUROC of X1 model: 0.95 -> 0.85Improved
apply IPW correction to PIDnaive complete-case analysis · Bitwise logic AND (MAR)Unique information 1: 0.44 -> 0.03Improved
apply IPW correction to PIDnaive complete-case analysis · Bitwise logic XOR (MAR)Complementary information: 0.62 -> 0.96Improved
apply IPW correction to PIDnaive complete-case analysis · Structural heart disease detectionUnique information of CXR: about 5% -> 1.8%Improved
apply IPW correction to PIDnaive complete-case analysis · Structural heart disease detectionShared information: 0.10 -> 0.48Improved
apply IPW correctionnaive complete-case analysis · UR-FUNNY (70% missingness)AUROC text: 0.61 -> 0.66Improved
apply IPW correctionnaive complete-case analysis · Hateful Memes (70% missingness)AUROC text: 0.68 -> 0.67No clear effect
correct both training and evaluation instead of one (ablation)IPW correction · simulationRMSE to oracle AUCImproved
apply IPW under MCARIPW correction · Bitwise logic (MCAR)AUROC / PIDNo clear effect
apply IPW under MNARIPW correction · Bitwise logic AND (MNAR)Unique information 1: 0.45 -> 0.45No clear effect
add CXR to ECGECG model · Structural heart disease detectionAUROC: 0.82 -> 0.83No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy