Research map / Video / Video-language understanding
Evaluating Reasoning Coherence in Video Generative Models with Text and Visual Hints
Yu Qi, Xinyi Xu, Ziyu Guo, Siyuan Ma, Renrui Zhang, Xinyan Chen, Ruichuan An, Ruofan Xing, Jiayi Zhang, Haojie Huang, Pheng-Ann Heng, Jonathan Tremblay and 1 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add text hints (ablation) | Veo3.1 · MME-CoF-Pro (16 categories) | Reasoning Score: +4.5 | Improved |
| add text hints (ablation) | Sora2 · MME-CoF-Pro (16 categories) | Reasoning Score: +7.6 | Improved |
| add text hints (ablation) | Cosmos-Predict2-14B · MME-CoF-Pro (16 categories) | Reasoning Score: +6.7 | Improved |
| add text hints (ablation) | Kling-v2.1 · MME-CoF-Pro (16 categories) | Reasoning Score: 13.8 (-0.4) | No clear effect |
| add text hints (ablation) | Sora2 · MME-CoF-Pro (16 categories) | Consistency Score: 60.5 (-2.4) | Hurt |
| add text hints (ablation) | 7 video models · 4D Dynamics Reasoning | Consistency Score: -1.2 to -15.6 | Hurt |
| add visual hints (ablation) | Veo3.1 · Visual Detail Reasoning (mini) | Reasoning Score: 27.8 (-13.0) | Hurt |
| add visual hints (ablation) | Seedance-1.0-fast · Visual Detail Reasoning (mini) | Reasoning Score: 27.8 (-13.0) | Hurt |
| add visual hints (ablation) | Veo3.1 · Embodied Reasoning (mini) | Reasoning Score: 60.8 (+22.9) | Improved |
| add visual hints (ablation) | Veo3.1 · GUI Reasoning (mini) | Reasoning Score: 27.5 (+17.5) | Improved |
| add visual hints (ablation) | 7 video models · MME-CoF-Pro-mini average | Reasoning Score: 36.9 (+10.8) | Mixed |
| add more text and visual hints in stages (ablation) | Sora2 · Frozen Lake | Reasoning Score: 0.23 | No clear effect |
Closest papers by meaning
- OpenCoF: Learning to Reason Through Video Generation (NeurIPS 2026 · similarity 0.92)
- CollabVR: Collaborative Video Reasoning with Vision-Language and Video Generation Models (NeurIPS 2026 · similarity 0.89)
- Thinking with Video: Video Generation as a Promising Multimodal Reasoning Paradigm (CVPR 2026 · similarity 0.89)
- When Thinking Drifts: Evidential Grounding for Robust Video Reasoning (NeurIPS 2025 · similarity 0.88)
- VideoReasonBench: Can MLLMs Perform Vision-Centric Complex Video Reasoning? (ICLR 2026 · similarity 0.88)
- VELOCITI: Benchmarking Video-Language Compositional Reasoning with Strict Entailment (CVPR 2025 · 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.
