AI Paper Atlas

Research map / Video / Video-language understanding

VideoZoomer: Reinforcement-Learned Temporal Focusing for Long Video Reasoning

Yang Ding, Xin Lai, Yizhen Zhang, Wei Li, Ruihang Chu, Yujiu Yang

ICLR 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 21 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
add RL stage after cold-start SFT (ablation)VideoZoomer · LongVideoReason-evalaccuracy: 63.3 -> 80.3Improved
add cold-start SFT initialization before RL (ablation)VideoZoomer · LongVideoReason-evalaccuracy: 59.6 -> 80.3Improved
add conditional tool-use bonus R_tool (ablation)VideoZoomer · VideoMME overallaccuracy: 62.5 -> 65.2Improved
add conditional tool-use bonus R_tool (ablation)VideoZoomer · VideoMMMUaccuracy: 53.8 -> 52.2Hurt
add reflection trajectories to cold-start data (ablation)VideoZoomer · VideoMME longaccuracy: 47.4 -> 55.8Improved
add reflection trajectories to cold-start data (ablation)VideoZoomer · VideoMMLU quizaccuracy: 70.1 -> 67.9Hurt
allow one tool call instead of none (ablation)VideoZoomer · LongVideoReason-evalaccuracy: 64.8 -> 79.9Improved
raise max tool calls from 2 to 4 (ablation)VideoZoomer · LongVideoReason-evalaccuracy: 80.2 -> 80.3No clear effect
replace uniform initial frames with TSPO-selected frames (ablation)VideoZoomer · MLVU devaccuracy: 68.8 -> 70.8Improved
distill cold-start data from Gemini 2.5 Pro instead of GPT-4o (ablation)VideoZoomer · VideoMME overallaccuracy: 61.6 -> 65.2Improved
double SFT data from about 11k to about 20k samples (ablation)VideoZoomer · LongVideoBench valaccuracy: 57.7 -> 55.6Hurt
train VideoZoomer with agentic zoom instead of base Qwen2.5-VLVideoZoomer · MLVU testaccuracy: 45.5 -> 55.8Improved

Later papers that compared with this method

2 later accepted papers report Video-Zoomer as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

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