1nm News
yc

Hemlock: The missing force signal in robotic training data

Hemlock · 1 points · 0 comments

We collect egocentric video data + force and contact dynamics derived from workers' muscular signals. Adding Hemlock's force data to human training demos improved pi0.5's success by 10x and Gemini Argon's by 2.7x on force-sensitive manipulation tasks.

新项目,适合观察早期产品方向。

查看原文 →

相关推荐

Sentient OS: On-device AI that knows you and takes over your work
yc · 17:00 10-27

Applied Kinetics: AI-native manufacturing for data center and energy projects
yc · 16:00 10-27

Rystic: Test your AI agents with local copies of Slack, Stripe, and Google Drive 🧪
yc · 23:00 10-26

Martini Bar: AI video production infrastructure
yc · 20:56 10-26

Hopper – Fast, open-source inference for real-time agents
yc · 19:00 10-26

AnyAlpha: The agentic trading platform
yc · 18:00 10-26