AI Paper Atlas

Research map / Video / Action recognition

VideoAgentTrek: Computer-Use Pretraining from Unlabeled Videos

Dunjie Lu, Yiheng Xu, Junli Wang, Haoyuan Wu, Xinyuan Wang, Zekun Wang, Junlin Yang, Hongjin SU, Jixuan Chen, Junda Chen, Yuchen Mao, Junyang Lin and 2 more

ICLR 2026 · Video › Action recognition · approach: Temporal action segmentation

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 stage 1 continued pretraining on mined video trajectories before SFT (ablation)Qwen2.5-VL-7B · OSWorld-Verifiedtask success rate (%): 9.3 -> 14.13 (20 steps); 9.27 -> 15.78 (50 steps)Improved
add stage 1 continued pretraining on mined video trajectories before SFT (ablation)Qwen2.5-VL-7B · AgentNetBenchstep success rate: 64.1% -> 69.3%Improved
scale video pretraining tokens from 0% to 50% to 100% (ablation)VideoAgentTrek · AgentNetBenchstep success rate: 64.1% -> 68.1% -> 69.3%Improved
scale video pretraining tokens from 0% to 50% to 100% (ablation)VideoAgentTrek · OSWorld-Verified SR@50task success rate (%): 9.3% -> 13.3% -> 15.7%Improved
increase step budget from 20 to 50 steps for model with video pretrainingVideoAgentTrek stage 1 + 2 · OSWorld-Verifiedtask success rate (%): 14.13 -> 15.78Improved
increase step budget from 20 to 50 steps for SFT-only model (ablation)SFT-only (stage 2 only) · OSWorld-Verifiedtask success rate (%): 9.3 -> 9.27No clear effect
add SFT on human demonstrations (without video pretraining) to base modelQwen2.5-VL-7B · AgentNetBenchstep success rate: 0.385 -> 0.641Improved
use ScreenFilter cursor-based filtering of candidate videosdata pipeline · corpus filteringGUI hours retained: 10,000 hours -> 7,377 hoursImproved

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