API Reference

This section contains the complete API reference for all modules in torchsparsegradutils.

• Core Operations
  • Sparse Matrix Multiplication
  • Indexed Matrix Multiplication
  • Sparse Linear Solvers
  • Sparse Least Squares
• Utility Functions
  • Iterative Solvers
  • Sparse Utilities
  • Random Sparse Tensor Generation
  • Statistical Distribution Helpers
• Probability Distributions
  • Sparse Multivariate Normal
  • Constraints
• Encoders
  • Pairwise Encoders
  • Pairwise Voxel Encoders
• Backend Integrations
  • CuPy Backend
  • JAX Backend

Core Functions

sparse_mm	Sparse–dense matrix multiplication with memory-efficient gradients.
gather_mm	Per-row indexed matrix multiplication.
segment_mm	Segmented matrix multiplication with variable-length segments.
sparse_triangular_solve	Sparse triangular solve with memory-efficient sparse gradients.
sparse_generic_solve	Sparse linear solve with iterative methods and sparse-aware gradients.
sparse_generic_lstsq	Sparse linear least squares with sparse-aware gradients.