brainevent.binary_jitnmm_p

brainevent.binary_jitnmm_p = <brainevent.XLACustomKernel object>

Low-level XLA custom-kernel primitive for binary_jitnmm.

This XLACustomKernel instance dispatches the binary (event-driven) JIT normal connectivity matrix-matrix multiplication operation to registered backends (numba, pallas), using runtime shape/dtype metadata provided by the high-level wrapper.

In this operation, the connectivity matrix has weights normally distributed with specified mean and standard deviation, and the input matrix is treated as binary events (spikes). Each column of the input is processed independently as an event vector.

Beyond backend dispatch, the primitive stores JAX transformation bindings (JVP, transpose, batching, and call registration) so the operation integrates correctly with jit, vmap, and autodiff.

Available backends can be queried with binary_jitnmm_p.available_backends(platform), and the default backend can be configured with binary_jitnmm_p.set_default(platform, backend).

See also

binary_jitnmm

High-level user-facing function wrapper.