UNO-Bench: A Unified Benchmark for Compositionality in Omni-Modal AI Models

UNO-Bench is a unified benchmark for exploring compositional relationships between uni-modal and omni-modal capabilities in AI models. The dataset was created by meituan-longcat and was last updated on December 4, 2025. It is accompanied by released evaluation scripts and a scoring model named UNO-Scorer-Qwen3-14B.

Use Cases

• Benchmarking the compositional reasoning of omni-modal models based on the described unified evaluation framework.
• Training or fine-tuning scoring models for multimodal tasks based on the released UNO-Scorer-Qwen3-14B model.
• Analyzing the performance gaps between uni-modal and omni-modal understanding based on the benchmark's structure.

Strengths

• Provides a unified evaluation framework (uno-eval) specifically for omni-modal benchmarks.
• Includes a released scoring model (UNO-Scorer-Qwen3-14B) for use with the benchmark.

Limitations

• Description metadata is limited; actual data quality, column definitions, and sample structure require manual inspection after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.

Provenance

Source

meituan-longcat on Hugging Face.

Collection Method

Likely constructed as a benchmark for AI model evaluation.

Time Range

Release activity noted in December 2025.

Freshness

Last updated 2025-12-04 07:53:07.

Geography

null