Research map / Vision-language and multimodal / Multimodal LLMs
Do Vision and Text Cues Exhibit Evidential Coupling? UFO: A Benchmark for Compositional Multimodal Reasoning in Unified Models
Zhongyu Yang, Dannong Xu, Yonghan Zhang, Kefan Chen, Xinyi Wang, Yang Xu, Wei Pang, Yingfang Yuan
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Introduces UFO, a benchmark probing whether visual content generated by unified foundation models serves as grounded evidence for subsequent multimodal reasoning.
Problem
Unclear whether visual content generated by Unified Foundation Models (UFMs) functions as grounded evidence for later reasoning or is merely auxiliary output; existing benchmarks evaluate generation and understanding separately.
Approach
A benchmark that evaluates interleaved generation-then-reasoning, testing whether generated visual cues are coupled as evidence to the textual reasoning that follows, rather than scoring generation and understanding in isolation.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | OmniGen2 · UFO MCQ State Reconstruction | accuracy: 51.52 -> 61.18 | Improved |
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | GPT-5.2 · UFO MCQ average | accuracy: 46.50 -> 52.12 | Improved |
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | Gemini-3.0-pro · UFO MCQ average | accuracy: 67.65 -> 66.13 | Hurt |
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | Gemini-3.0-pro · UFO open questions average | accuracy: 63.69 -> 66.42 | Improved |
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | Bagel · UFO MCQ average | accuracy: 42.54 -> 41.70 | Hurt |
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | UniWorld-V1 · UFO open questions average | accuracy: 28.11 -> 26.19 | Hurt |
| add generated visual and textual cues (Joint schedule) over Direct (ablation) | Ovis-U1 · UFO open questions average | accuracy: 24.51 -> 22.48 | Hurt |
| add visual cue to textual cue (Joint vs Textual) (ablation) | UniCoT · UFO MCQ State Augmentation | accuracy: 39.65 -> 37.19 | Hurt |
| add generated cues (Joint) over Direct (ablation) | Gemini-3.0-pro · UFO MCQ State Augmentation | accuracy: 52.32 -> 45.42 | Hurt |
| switch from multiple-choice to open-ended answering | UniPic1 · UFO State Determination | accuracy: 25.91 -> 0.96 | Hurt |
| compare visual to textual cue generation | Janus-Pro 1B · UFO State Determination cue accuracy | cue accuracy: 22.13 -> 2.34 | Hurt |
| use Chain-of-Thought judging instead of Binary (ablation) | Gemini-3-Flash judge · human agreement (average kappa) | Cohen kappa: 23.9 -> 30.1 | Improved |
Experimental setup
Benchmarks: UFO
Closest papers by meaning
- ROVER: Benchmarking Reciprocal Cross-Modal Reasoning for Omnimodal Generation (ICLR 2026 · similarity 0.90)
- Compose and Fuse: Revisiting the Foundational Bottlenecks in Multimodal Reasoning (ICLR 2026 · similarity 0.89)
- Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens (CVPR 2026 · similarity 0.89)
- Look on Demand: A Cognitive Scheduling Framework for Visual Evidence Acquisition in Multimodal Reasoning (ICML 2026 · similarity 0.88)
- LENS: Multi-level Evaluation of Multimodal Reasoning with Large Language Models (ICLR 2026 · similarity 0.88)
- ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning (ICLR 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
