AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

ComputerRL: Scaling End-to-End Online Reinforcement Learning for Computer Use Agents

Hanyu Lai, Xiao Liu, Yanxiao Zhao, Han Xu, Hanchen Zhang, Bohao Jing, Yanyu Ren, Shuntian Yao, Yuxiao Dong, Jie Tang

ICLR 2026 · Reinforcement learning › Visual and representation RL · approach: Scalable RL systems and environments · 56 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 API-GUI action space instead of GUI only (ablation)GPT-4o agent · OSWorldaverage success: 11.2 -> 26.2Improved
add behavior cloning cold start (ablation)Qwen2.5-14B · OSWorldaverage success: 15.2 -> 31.9Improved
add RL phase 1 (ablation)Qwen2.5-14B · OSWorldaverage success: 31.9 -> 42.0Improved
add Entropulse (SFT on successful rollouts) after RL1 (ablation)Qwen2.5-14B · OSWorldaverage success: 42.0 -> 41.5No clear effect
add RL phase 2 after Entropulse (ablation)Qwen2.5-14B · OSWorldaverage success: 41.5 -> 45.8Improved
add Entropulse before second RL phase vs plain continued RL (ablation)ComputerRL · training reward and entropyrewardImproved
raise clipping threshold (DAPO-style) to slow entropy decline (ablation)ComputerRL RL · entropy / learningpolicy improvementHurt
use RL-trained ComputerRL 9B vs strongest proprietary baselineGLM-4.1V-9B-Thinking · OSWorldsuccess rate: 42.9 -> 48.9Improved
train with ComputerRL on GLM-4.1V vs GLM-4-9B backboneAutoGLM-OS · OSWorldsuccess rate: 48.1 -> 48.9No clear effect
use ComputerRL agent vs OpenAI o3 on PowerPoint tasksGLM-4.1V-9B-Thinking · OfficeWorld PPTsuccess: 41.7 -> 41.7No clear effect

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