Open Robot
Intelligence
XRollout is a passionate community pushing the boundaries of robot learning. From VLA to Diffusion Policy, we pursue extreme generalization — one model that adapts to any robot, any task, any environment.
from xrollout import VLAConfig, Trainer
from xrollout.models import DiffusionPolicy
# 配置 VLA 模型
config = VLAConfig(
vision_encoder="siglip-so400m",
llm="qwen2-7b",
action_dim=7,
horizon=16
)
# 初始化训练器
trainer = Trainer(
config=config,
dataset="xrollout/robotics-v1",
policy=DiffusionPolicy(config)
)
# 开始训练
trainer.train(
batch_size=64,
num_epochs=100,
mixed_precision=True
)Essays & Philosophy
Browse all →Our Mission
Read more about why we started this project and how we want to democratize robot learning for everyone.
The Art of Deliberate Practice
Our complete learning philosophy — four pillars of robot acquisition and the hierarchical data pyramid.
Why Language: Human Brain Perspective
From prefrontal cortex vs basal ganglia to why both vision and language are indispensable in VLA.
XRollout 项目已建成功能
从每日论文订阅，到完整 SLAM/重建/标定工具链，再到社区文章分享平台，一览 XRollout 已经建成的一切。
Fast-WAM 深度解读
世界动作模型真的需要在测试时想象未来吗？清华团队发现视频预测的主要价值来自训练阶段而非测试阶段，速度提升 4 倍。
MEM 多尺度具身记忆 深度解读
Physical Intelligence 最新研究：让机器人拥有长短时记忆，通过混合模态架构解决长达 15 分钟的复杂操作任务。
Share Your Vision — Help Shape the Future of Robotics
We're collecting community opinions on robot form factors, applications, and how robots should coexist with humans. Your voice matters.
Discover
Browse all →A cost-effective distributed human data collection pipeline for imitation learning. Collect thousands of teleoperation demonstrations from crowd workers via browser.
The Agent AI Skill Program - Learn how to build AI agents that can acquire new skills through interactive demonstration and practice.