AI Paper Atlas

Research map / Video / Video-language understanding

Towards Video Thinking Test: A Holistic Benchmark for Advanced Video Reasoning and Understanding

Yuanhan Zhang, Yunice Chew, Yuhao Dong, Aria Leo, Bo Hu, Ziwei Liu

ICCV 2025 · Video › Video-language understanding · approach: Long video understanding · 9 citations (OpenAlex)

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 changedWhereMeasuredEffect
use a larger model (72B instead of 7B) (ablation)Qwen2.5-VL · Video-TTrobustness score: 14.4 -> 22.2Improved
use a larger model (72B instead of 7B) (ablation)LLaVA-Video · Video-TTrobustness score: 13.7 -> 19.7Improved
scale InternVL-2.5 from 8B to 38B (ablation)InternVL-2.5 · Video-TTrobustness score: 10.9 -> 11.1No clear effect
use video frames instead of language-only blind input (ablation)GPT-4o · Video-TTprimary open-ended accuracy: 8.5 -> 36.6Improved
give correct cues in the question (correctly-led) instead of primary questionGPT-4o (blind) · Video-TTaccuracy: 8.5 -> 58.9Improved
add Chain-of-Thought prompt (Let's think step by step) (ablation)video LLMs · Video-TT wrongly-ledrelative accuracy change: about 6.8% relative increaseImproved
add Chain-of-Thought prompt (ablation)video LLMs · Video-TT multiple-choiceaccuracyNo clear effect
add audio transcript (ablation)GPT-4o · Video-TT multiple-choiceaccuracyNo clear effect
add audio transcript (ablation)GPT-4o · Video-TTrobustness score: almost 15% relative gainImproved
increase the number of input frames beyond 8 (ablation)video LLMs · Video-TTaccuracyNo clear effect
increase the number of input frames up to 64 (ablation)human annotators · Video-TTaccuracyImproved
use multiple-choice instead of primary open-ended formatLLaVA-Video-72B · Video-TTaccuracy: 24.4 -> 47.5Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy