nowflake
Snowflake: Scaling GNNs to high-dimensional continuous control via parameter freezing Charlie Blake
GNN s deteriorating rapidly as the number of sensors and actuators grows.
Technology:
References [1 ]
Neural learning in structured parameter spaces-natural riemannian gradient. Reward propagation using graph convolutional networks. Attention is all you need. Our work addresses the problem of scaling GNNs for simulated locomotion control. As we only ever conduct rollouts of the policy in simulation, our agent is not trained with any safety constraints in mind, which would likely be a requirement for real-world applications.
Country:
- Asia > Middle East > Jordan (0.05)
- Europe > Romania > Sud - Muntenia Development Region > Giurgiu County > Giurgiu (0.04)
- Europe > Italy > Calabria > Catanzaro Province > Catanzaro (0.04)
- Asia > India (0.04)
Technology:
Snowflake: Scaling GNNs to high-dimensional continuous control via parameter freezing Charlie Blake
GNN s deteriorating rapidly as the number of sensors and actuators grows.
Technology:
Country:
- Asia > Middle East > Jordan (0.05)
- Europe > Romania > Sud - Muntenia Development Region > Giurgiu County > Giurgiu (0.04)
- Europe > Italy > Calabria > Catanzaro Province > Catanzaro (0.04)
- Asia > India (0.04)
Technology: