AI Paper Atlas

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

Same Content, Different Answers: Cross-Modal Inconsistency in MLLMs

Angela van Sprang, Laurens Samson, Ana Lucic, Erman Acar, Sennay Ghebreab, Yuki M. Asano

CVPR 2026 · Vision-language and multimodal › Cross-modal learning · approach: Contrastive vision-language alignment · 9 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
present question as text instead of imageGPT-4o-mini · MMLUaccuracy: 77.0 -> 84.4Improved
present question as text instead of imagePhi-4 · MMLUaccuracy: 31.4 -> 47.0Improved
present question as text instead of imagePhi-4 · SoEBenchaccuracy: 17.3 -> 13.3Hurt
raise rendering resolution from DPI 50 to 200 (ablation)Haiku-4.5 · REST+ (all questions)RER: 53.6 -> 76.3Improved
raise rendering resolution from DPI 50 to 200 (ablation)GPT-5-mini · REST+ (all questions)RER: 71.5 -> 83.6Improved
raise rendering resolution from DPI 50 to 200 (ablation)Gemini-2.5 Flash Lite · REST+ (all questions)RER: 70.6 -> 70.9No clear effect
change font family (DejaVu Sans, Courier New, Cursive) (ablation)MLLMs · REST+image accuracy: within 2% absolute differenceNo clear effect
use coloured (red or yellow) text instead of black (ablation)MLLMs · REST+ (DPI@200)image accuracy: more than 5% relative improvementImproved
OCR first then solve instead of direct image processing (ablation)Deepseek-Small · MMLUaccuracy: -13.1Hurt
OCR first then solve instead of direct image processing (ablation)Qwen-2.5-32B · SoEBenchaccuracy: +10.1Improved
OCR first then solve instead of direct image processing (ablation)Phi-4 · AI2-ARCaccuracy: +9.0Improved
OCR first then solve instead of direct image processing (ablation)MLLMs · RESTaccuracyMixed

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