IL for manipulation

Robot Learning Collective: VLA

(outdated)

RL for locomotion

• Finish Hugging Face RL course (practical ‘coding’ part)
• Re-implement PPO
• Train locomotion in sim (maybe simplify the code)
• Re-implement end-to-end navigation + locomotion paper

LLMs

• Read the reasoning paper from DeepSeek (R1)
• Finish Karpathy's “course”
• …?

Papers to go over

• https://generalistai.com/blog.html ✅
• SmalVLA ✅
• https://toyotaresearchinstitute.github.io/lbm1/ ✅
• Notebook with DP and Flow matching https://youtu.be/6oZe_tKE3YA?si=H9hvmhHRGeiM_V9h
• https://dit-policy.github.io/