Event
TD-MPC2: Reproducing Latent-Space World Model Control in Practice
1 reports · 1 sources
Using the official pretrained weights, the author drove a quadruped canine model to complete real running inference on a local CPU, and reproduced the mathematical construction of TD-MPC2 item by item, including the latent-space world model, the consistency learning objective, symlog two-point hot encoding, and MPPI-style sampling-based planning. The author recorded empirically observed phenomena such as planning distribution contraction and the gap between estimates and realized outcomes.
Report timeline
- TD-MPC2: Model-Predictive Reinforcement Learning Control Practice with Latent World Models
The author uses official pretrained weights to drive a quadruped canine model through real running inference on a local CPU, reconstructing item by item the mathematical construction of the latent world model, the consistency learning objective, symlog two-hot encoding, and MPPI-style sampling-based planning, and records observed phenomena such as planning distribution contraction and the gap between estimates and realized outcomes.
Guyuehome ·