Online-Mind-2Web
Emerging5papers using it
2025first seen
The 'Online-Mind-2Web' dataset/benchmark contains a collection of interactive tasks and environments designed to evaluate the performance of computer-use agents (CUAs) in executing tasks through vision-language models.
Papers using Online-Mind-2Web (5)
- OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web AgentsOrchard: An Open-Source Agentic Modeling FrameworkFrom Off-Policy to On-Policy: Enhancing GUI Agents via Bi-level Expert-to-Policy AssimilationUI-TARS-2 Technical Report: Advancing GUI Agent with Multi-Turn Reinforcement LearningTGPO: Tree-Guided Preference Optimization for Robust Web Agent Reinforcement Learning