Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.18915 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908619637784576 |
|---|---|
| author | Chen, Chen Hu, ZeYang Chen, Fengjiao Ma, Liya Liu, Jiaxing Li, Xiaoyu Wang, Ziwen Cao, Xuezhi Cai, Xunliang |
| author_facet | Chen, Chen Hu, ZeYang Chen, Fengjiao Ma, Liya Liu, Jiaxing Li, Xiaoyu Wang, Ziwen Cao, Xuezhi Cai, Xunliang |
| contents | Multimodal Large Languages models have been progressing from uni-modal understanding toward unifying visual, audio and language modalities, collectively termed omni models. However, the correlation between uni-modal and omni-modal remains unclear, which requires comprehensive evaluation to drive omni model's intelligence evolution. In this work, we introduce a novel, high-quality, and UNified Omni model benchmark, UNO-Bench. This benchmark is designed to effectively evaluate both UNi-modal and Omni-modal capabilities under a unified ability taxonomy, spanning 44 task types and 5 modality combinations. It includes 1250 human curated samples for omni-modal with 98% cross-modality solvability, and 2480 enhanced uni-modal samples. The human-generated dataset is well-suited to real-world scenarios, particularly within the Chinese context, whereas the automatically compressed dataset offers a 90% increase in speed and maintains 98% consistency across 18 public benchmarks. In addition to traditional multi-choice questions, we propose an innovative multi-step open-ended question format to assess complex reasoning. A general scoring model is incorporated, supporting 6 question types for automated evaluation with 95% accuracy. Experimental result shows the Compositional Law between omni-modal and uni-modal performance and the omni-modal capability manifests as a bottleneck effect on weak models, while exhibiting synergistic promotion on strong models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_18915 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UNO-Bench: A Unified Benchmark for Exploring the Compositional Law Between Uni-modal and Omni-modal in Omni Models Chen, Chen Hu, ZeYang Chen, Fengjiao Ma, Liya Liu, Jiaxing Li, Xiaoyu Wang, Ziwen Cao, Xuezhi Cai, Xunliang Computation and Language Artificial Intelligence I.2.7 Multimodal Large Languages models have been progressing from uni-modal understanding toward unifying visual, audio and language modalities, collectively termed omni models. However, the correlation between uni-modal and omni-modal remains unclear, which requires comprehensive evaluation to drive omni model's intelligence evolution. In this work, we introduce a novel, high-quality, and UNified Omni model benchmark, UNO-Bench. This benchmark is designed to effectively evaluate both UNi-modal and Omni-modal capabilities under a unified ability taxonomy, spanning 44 task types and 5 modality combinations. It includes 1250 human curated samples for omni-modal with 98% cross-modality solvability, and 2480 enhanced uni-modal samples. The human-generated dataset is well-suited to real-world scenarios, particularly within the Chinese context, whereas the automatically compressed dataset offers a 90% increase in speed and maintains 98% consistency across 18 public benchmarks. In addition to traditional multi-choice questions, we propose an innovative multi-step open-ended question format to assess complex reasoning. A general scoring model is incorporated, supporting 6 question types for automated evaluation with 95% accuracy. Experimental result shows the Compositional Law between omni-modal and uni-modal performance and the omni-modal capability manifests as a bottleneck effect on weak models, while exhibiting synergistic promotion on strong models. |
| title | UNO-Bench: A Unified Benchmark for Exploring the Compositional Law Between Uni-modal and Omni-modal in Omni Models |
| topic | Computation and Language Artificial Intelligence I.2.7 |
| url | https://arxiv.org/abs/2510.18915 |