What's Left Unsaid? Detecting and Correcting Misleading Omissions in Multimodal News Previews

TL;DR

This paper introduces a new benchmark and method for detecting and correcting misleading omissions in multimodal news previews, improving understanding and reducing interpretation drift.

Contribution

It develops a multi-stage pipeline and OMGuard system that enhance detection accuracy and enable effective correction of misleading omissions in social-media news previews.

Findings

OMGuard improves detection accuracy to match 235B LVLMs from 8B models.

Misleadingness often results from local narrative shifts like missing background.

Visual interventions are crucial for image-driven misleading content correction.

Abstract

Even when factually correct, social-media news previews (image-headline pairs) can induce interpretation drift: by selectively omitting crucial context, they lead readers to form judgments that diverge from what the full article supports. This covert harm is subtler than explicit misinformation, yet remains underexplored. To address this gap, we develop a multi-stage pipeline that simulates preview-based and context-based understanding, enabling construction of the MM-Misleading benchmark. Using MM-Misleading, we systematically evaluate open-source LVLMs and uncover pronounced blind spots in omission-based misleadingness detection. We further propose OMGuard, which combines (1) Interpretation-Aware Fine-Tuning for misleadingness detection and (2) Rationale-Guided Misleading Content Correction, where explicit rationales guide headline rewriting to reduce misleading impressions. Experiments show that OMGuard lifts an 8B model's detection accuracy to the level of a 235B LVLM while delivering markedly stronger end-to-end correction. Further analysis shows that misleadingness usually arises from local narrative shifts, such as missing background, instead of global frame changes, and identifies image-driven cases where text-only correction fails, underscoring the need for visual interventions.