Robot Learning Notes
Technical visual notes for papers and model architectures I want to remember quickly.
π₀.₇: A Steerable Generalist Robotic Foundation Model with Emergent Capabilities
RL Token: Bootstrapping Online RL with Vision-Language-Action Models
Dyna-2: A 1-Million-Hour Scaling Law for World-Action Models
MolmoAct2: Action Reasoning Models for Real-world Deployment
Action-to-Action Flow Matching
MEM: Multi-Scale Embodied Memory for Vision Language Action Models
FAST: Frequency-space Action Sequence Tokenization
Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisons
π₀.₅: A Vision-Language-Action Model with Open-World Generalization
π₀: A Vision-Language-Action Flow Model for General Robot Control
MolmoMotion: Forecasting Point Trajectories in 3D with Language Instruction
ACT: Action Chunking with Transformers