Skip to content

Same Semantics, Different Outcome: On the Modality Robustness of Multimodal LLMs under Knowledge Conflict

Sep 2026 · 1 citation · 26 references
Computer Science

TL;DR

Modality robustness under knowledge conflict is studied across 13 MLLMs and two datasets, and it is found that instability has practical consequences: it degrades performance in multimodal RAG and can be exploited by adversarial attacks.

Abstract

Multimodal large language models (MLLMs) are increasingly provided with contextual evidence in heterogeneous forms: as a text passage, as a rendered image of the same passage, or as both together. However, it remains unclear how consistently these surface forms are processed, especially when the evidence conflicts with the model's parametric knowledge. We study modality robustness under knowledge conflict across 13 MLLMs and two datasets, and find them far from robust. (1) Contrary to common belief, models favor a context that contradicts parametric knowledge more readily in image form than in text form; (2) when a contradicting text and image are presented together, the preferred modality is essentially arbitrary, varying with input order, model, and dataset. We further demonstrate that this instability has practical consequences: it degrades performance in multimodal RAG and can be exploited by adversarial attacks. To alleviate this brittleness, we examine several simple techniques---prompting, steering, supervised fine-tuning (SFT), and direct preference optimization; the majority prove ineffective, whereas SFT achieves moderate success. We therefore call for greater awareness of this inconsistency and argue that it is fundamental, demanding attention at multiple training stages.

View source

Similar papers

Open access Sep 2026

Quantifying and reducing the modality gap in both contrastive and generative multimodal vision–language models

Dual-encoder vision-language models (VLMs) allow to embed texts and images in a shared representation space. However, it has been shown that these models are subject to a “modality gap phenomenon” meaning there exists a clear separation between the embeddings from one modality and another in the embedding space. While...

François Role, Sébastien Meyer, Victor Amblard · 0 citations
Open access Sep 2026

Forbid Your Attention: Fooling Multimodal Large Language Models by Selectively Removing Intrinsic Focus in Spectral Domain

This paper proposes a novel phase-aware adversarial attack framework that explicitly restricts adversarial perturbations to structure-relevant phase regions to suppress the MLLM' focus for effective and imperceptible attacks.

Dai-Zong Liu, Junhao Dong, Zhi-Yuan Ma et al. · 1 citation
#artificial intelligence Preprint Sep 2026

Layers, Sinks, and Scaling: Adaptive Evidence Selection for Multimodal Large Language Models

Adaptive Relevance-guided Evidence Allocation (AREA) is introduced, a training-free inference-time method that formulates evidence highlighting as adaptive allocation and establishes the best performance among training-free highlighting methods.

Zhen-Bin Wang, Lei Zhang, Li-Tuan Wang et al. · 0 citations
#computer vision Sep 2026

FORUM: Frozen Outputs Reconciled Using Model Agreement for Visual Grounding

ForUM is presented, a training-free test-time fusion of frozen MLLMs guided by two fixed geometric rules: agreement-based selection keeps the region supported by the most distinct models, and medoid localization returns an actual member box instead of a coordinate average, so one loose prediction cannot shift the answe...

Taiyo Sato, Takamasa Sanda, Keisuke Maeda et al. · 0 citations
Preprint Aug 2026

Beyond Language Priors: Diagnosing and Fixing Visual-Origin Hallucinations in Multimodal LLM

Adversarial Contrastive Fine-Tuning (ACFT) uses an Adversarial Hallucination Attribute Flipping procedure, involving minimal, targeted adversarial perturbations that flip an image's hallucination attribute, to construct perfectly aligned positive-negative pairs, which are then used for contrastive fine- tuning.

Pei-Yang Xu, Xiaopei Zhu, Jun Zhu et al. · 0 citations

Related blog posts

MIT News · Artificial Intelligence Nov 17, 2026

Learning the language of machine learning

By studying the computational foundations of language learning, Associate Professor Jacob Andreas builds intelligent systems that can communicate effectively with humans.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.