Paper Detail

Beyond Text-Dominance: Understanding Modality Preference of Omni-modal Large Language Models

Xinru Yan, Boxi Cao, Yaojie Lu, Hongyu Lin, Weixiang Zhou, Le Sun, Xianpei Han

huggingface Score 10.5

Published 2026-04-18 · First seen 2026-04-21

General AI

Abstract

Native Omni-modal Large Language Models (OLLMs) have shifted from pipeline architectures to unified representation spaces. However, this native integration gives rise to a critical yet underexplored phenomenon: modality preference. To bridge this gap, we first systematically quantify modality preference of OLLMs using a newly-curated conflict-based benchmark and the modality selection rate metric. Our evaluation of ten representative OLLMs reveals a notable paradigm shift: unlike the ``text-dominance'' of traditional VLMs, most OLLMs exhibit a pronounced visual preference. To further understand the underlying mechanism, we conduct layer-wise probing and demonstrate that such modality preference is not static but emerges progressively in the mid-to-late layers. Building upon these insights, we leverage these internal signals to diagnose cross-modal hallucinations, achieving competitive performance across three downstream multi-modal benchmarks without task-specific data. Our work provides both a mechanistic understanding and a practical tool for building more trustworthy OLLMs. Our code and related resources are publicly available at: https://github.com/icip-cas/OmniPreference

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@misc{yan2026beyond,
  title = {Beyond Text-Dominance: Understanding Modality Preference of Omni-modal Large Language Models},
  author = {Xinru Yan and Boxi Cao and Yaojie Lu and Hongyu Lin and Weixiang Zhou and Le Sun and Xianpei Han},
  year = {2026},
  abstract = {Native Omni-modal Large Language Models (OLLMs) have shifted from pipeline architectures to unified representation spaces. However, this native integration gives rise to a critical yet underexplored phenomenon: modality preference. To bridge this gap, we first systematically quantify modality preference of OLLMs using a newly-curated conflict-based benchmark and the modality selection rate metric. Our evaluation of ten representative OLLMs reveals a notable paradigm shift: unlike the ``text-domi},
  url = {https://huggingface.co/papers/2604.16902},
  keywords = {omni-modal large language models, modality preference, conflict-based benchmark, modality selection rate, cross-modal hallucinations, layer-wise probing, code available, huggingface daily},
  eprint = {2604.16902},
  archiveprefix = {arXiv},
}

Metadata

{}