Core Concepts
CSMPNs integrate Clifford group-equivariant layers with simplicial message passing for efficient geometric tasks.
Abstract
CSMPNs combine Clifford algebra with simplicial message passing for E(n)-equivariant operations.
The method efficiently shares parameters across different simplex dimensions, enhancing performance on geometric tasks.
Experimental results show superior performance over existing methods in various domains and dimensionalities.
Quotes
"We introduce Clifford Group Equivariant Simplicial Message Passing Networks."
"Our method integrates the expressivity of Clifford group-equivariant layers with simplicial message passing."