Share your thoughts, 1 month free Claude Pro on usSee more

Home/Benchmarks

Pick Apple on Real-world

95Success Rate

WORLD4RL

Updated 27d ago

Evaluation Results

Method	Links
WORLD4RL 2025.09		95
DP 2025.09		75
BC (Base) 2025.09		40

SOTA Paper

WORLD4RL

World4RL: Diffusion World Models for Policy Refinement with Reinforcement Learning for Robotic Manipulation

Dataset

Real-world

Follow for update

@wizwand_team Discord

Related Benchmarks

Video Prediction on Meta-World (Policy rollouts)Video Prediction on Meta-World Random rollouts Pick bread out on Real-world Press button on Real-world Put bread in on Real-world

© 2026 wizwand

Blog Contact Changelog Swarm

Privacy Policy Terms of Service FAQs Swarm Docs