Research map / Video / Video-language understanding
A Very Big Video Reasoning Suite
Maijunxian Wang, Ruisi Wang, Juyi Lin, Ran Ji, Thaddäus Wiedemer, Qingying Gao, Dezhi Luo, Yaoyao Qian, Lianyu Huang, Zelong Hong, Jiahui Ge, Qianli Ma and 43 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 |
|---|---|---|---|
| fine-tune Wan2.2-I2V-A14B with LoRA on VBVR-Dataset (500K samples) | Wan2.2-I2V-A14B · VBVR-Bench overall | score: 0.371 -> 0.685 | Improved |
| scale training data from 0K to 400K samples (ablation) | VBVR-Wan2.2 · VBVR-Bench in-domain | score: 0.412 -> 0.771 | Improved |
| increase training data beyond 400K samples (ablation) | VBVR-Wan2.2 · VBVR-Bench in-domain | score: 0.771 -> 0.760 | No clear effect |
| scale training data from 0K to 500K samples (ablation) | VBVR-Wan2.2 · VBVR-Bench out-of-domain | score: 0.329 -> 0.610 | Improved |
| increase training data from 200K to 500K samples (ablation) | VBVR-Wan2.2 · VBVR-Bench overall | score: 0.689 -> 0.685 | No clear effect |
| apply the same VBVR fine-tuning to a different backbone (LTX-2.3) (ablation) | LTX-2.3 · VBVR-Bench overall | score: 0.516 overall | Improved |
| LoRA fine-tune on VBVR-Dataset | Wan2.2-I2V-A14B · VBench-I2V camera motion consistency | score: 0.5444 -> 0.6592 | Improved |
| LoRA fine-tune on VBVR-Dataset | Wan2.2-I2V-A14B · VBench-I2V dynamic degree | score: 0.5285 -> 0.4106 | Hurt |
| LoRA fine-tune on VBVR-Dataset | Wan2.2-I2V-A14B · VBench-I2V total score | score: 0.8816 -> 0.8835 | No clear effect |
Closest papers by meaning
- CollabVR: Collaborative Video Reasoning with Vision-Language and Video Generation Models (NeurIPS 2026 · similarity 0.92)
- Video-of-Thought: Step-by-Step Video Reasoning from Perception to Cognition (ICML 2024 · similarity 0.92)
- VideoReasonBench: Can MLLMs Perform Vision-Centric Complex Video Reasoning? (ICLR 2026 · similarity 0.91)
- VideoEspresso: A Large-Scale Chain-of-Thought Dataset for Fine-Grained Video Reasoning via Core Frame Selection (CVPR 2025 · similarity 0.91)
- Look, Remember and Reason: Grounded Reasoning in Videos with Language Models (ICLR 2024 · similarity 0.91)
- SciVideoBench: Benchmarking Scientific Video Reasoning in Large Multimodal Models (ICML 2026 · similarity 0.91)
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.
