rpzhou

[rpzhou @rpzhou]

📚 Paper：https://t.co/n4Q3iRjrk5
🌐 Project Page：https://t.co/UOlIuI2QYU
🤗 Hugging Face：https://t.co/KkAu4QOe9X
🔗 Code：https://t.co/iPBOKTbYU2

[rpzhou @rpzhou]

Delighted to see Simple-WAM top the Hugging Face "Daily Papers" list. Although rankings fluctuate in real-time, we would like to take this moment to briefly introduce the latest work from our LeapLab team.

[rpzhou @rpzhou]

🔍 Evaluating World Action Model (WAM) generalization: Environment Perturbation, Data Efficiency, Unseen Tasks. Key Findings: keep future representations at inference, not generate clear futures.
⚡ Simple-WAM: one forward pass, sim+real generalization, 3.8× faster.