Equivariant Graph Neural Network Vault (equivariant-graph-neural-network-vault)

Comprehensive scientific research vault housing algorithms, Lie-group symmetry transformations, symplectic integrators, and equivariant coordinate message passing implementations for $E(n)$ Equivariant Graph Neural Networks (EGNN).

Vault Repository Structure

equivariant-graph-neural-network-vault/
β”œβ”€β”€ src/
β”‚   β”œβ”€β”€ equivariant_layers.py      # E(n) Equivariant message passing and coordinate updates
β”‚   β”œβ”€β”€ symplectic_integrator.py   # Symplectic Velocity Verlet solver for Hamiltonian systems
β”‚   └── lie_group_transforms.py    # SO(3) rotation matrices and Rodrigues group generators
β”œβ”€β”€ benchmarks/
β”‚   └── eval_equivariance_suite.py # Rotational and translational equivariance verification
β”œβ”€β”€ MANIFEST.json                  # Cryptographic SHA-256 integrity manifest
└── README.md                      # Mathematical formulations and documentation

Equations

  • $E(n)$ Equivariant Message Passing: $$m_{ij} = \phi_e(h_i^l, h_j^l, |x_i^l - x_j^l|^2, a_{ij})$$ $$x_i^{l+1} = x_i^l + \frac{1}{|\mathcal{N}(i)|} \sum_{j \in \mathcal{N}(i)} (x_i^l - x_j^l) \phi_x(m_{ij})$$ $$h_i^{l+1} = h_i^l + \phi_h(h_i^l, \sum_{j \in \mathcal{N}(i)} m_{ij})$$

  • Equivariance Guarantee: $$\mathbf{X}^{l+1}(R \mathbf{X} + \mathbf{t}) = R \mathbf{X}^{l+1}(\mathbf{X}) + \mathbf{t} \quad \forall R \in O(3), \ \mathbf{t} \in \mathbb{R}^3$$

Reproducibility

python benchmarks/eval_equivariance_suite.py

Developed by Ramprasad Atluri (At1uri).

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support