AI Paper Atlas

Research map / Video / Video-language understanding

Video Streaming Thinking: VideoLLMs Can Watch and Think Simultaneously

Yiran Guan, Liang Yin, Dingkang Liang, Jianzhong Ju, Zhenbo Luo, Jian Luan, Yuliang Liu, Xiang Bai

ECCV 2026 · Video › Video-language understanding · approach: Long video understanding · 22 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
mix 30K VST streaming-thought data with 20K LLaVA-Vid in SFT instead of 50K LLaVA-Vid (ablation)Qwen2.5-VL-7B · OVO-Bench overallaccuracy: 52.3 -> 57.1Improved
add VST data to SFT mix (ablation)Qwen2.5-VL-7B · VideoMMEaccuracy: 61.8 -> 63.1No clear effect
add VST-SFT to base model (ablation)Qwen2.5-VL-7B · OVO-Bench Backwardaccuracy: 47.5 -> 56.7Improved
add VST-RL to base model (ablation)Qwen2.5-VL-7B · OVO-Bench Forwardaccuracy: 41.9 -> 54.6Improved
add RL on top of SFT (ablation)VST-7B · OVO-Bench overallaccuracy: 57.4 -> 59.3Improved
add RL on top of SFT (ablation)VST-7B · VideoMMEaccuracy: 63.0 -> 64.9Improved
apply VST training to 3B base (ablation)Qwen2.5-VL-3B · StreamingBenchaccuracy: 67.8 -> 75.5Improved
apply VST training to 7B base (ablation)Qwen2.5-VL-7B · StreamingBenchaccuracy: 71.7 -> 79.5Improved
apply VST training to 32B base (ablation)Qwen2.5-VL-32B · StreamingBenchaccuracy: 71.5 -> 80.7Improved
apply VST training to 7B base (ablation)Qwen2.5-VL-7B · VideoHolmesaccuracy: 32.9 -> 41.9Improved
increase max streaming thinking steps from 1 to 16 (ablation)VST-7B · OVO-Bench Backwardaccuracy: 53.3 -> 57.5Improved
increase max streaming thinking steps beyond 4 (ablation)VST-7B · OVO-Bench Real-Time and ForwardaccuracyNo clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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