Crate sc_neurocore_engine

Source

Modules§

attention: Stochastic Attention
bitstream: Bitstream Operations
brunel: Fused Brunel balanced network simulation in fixed-point Q8.8 arithmetic.
connectome: Biologically plausible connectivity generators.
conv: SC 2D convolutional layer using probability-domain multiplication.
cordiv: CORDIV: stochastic computing divider.
cortical_column: 5-population cortical column (Douglas & Martin 2004). L4 (thalamic) → L2/3 exc ↔ L2/3 inh → L5 (output) → L6 (feedback → L4).
encoder: Stochastic Encoders
fault: Hardware fault injection for robustness testing.
fusion: Multi-modal fusion layer using stochastic multiplexing.
grad
graph: Stochastic Graph Layer
ir: SC Compute Graph IR
layer: Dense Stochastic Layer
network_runner: High-performance network simulation backend.
neuron: Neuron Models
neurons
phi: Barrett & Seth 2011 geometric Phi under Gaussian assumption. Phi* = MI(past; future) - max_partition sum MI(past_k; future_k)
predictive_coding: Predictive coding via XOR + popcount in packed bitstream domain.
pyo3_neurons: PyO3 wrappers for all neuron models.
rall_dendrite: Compartmental dendritic tree with Rall’s 3/2 power rule. Distal → proximal propagation with inter-compartment coupling.
recorder: Spike recording and statistics.
recurrent: SC recurrent/reservoir layer (echo state network).
scpn
simd: SIMD Popcount Dispatch
sobol: Sobol quasi-random bitstream generator.
synapses: Stochastic STDP Synapse

Structs§

Functions§

__pyfunction_batch_encode 🔒 ^⚠
__pyfunction_batch_encode_numpy 🔒 ^⚠
__pyfunction_batch_lif_run 🔒 ^⚠
__pyfunction_batch_lif_run_multi 🔒 ^⚠
__pyfunction_batch_lif_run_varying 🔒 ^⚠
__pyfunction_ir_emit_sv 🔒 ^⚠
__pyfunction_ir_parse 🔒 ^⚠
__pyfunction_ir_print 🔒 ^⚠
__pyfunction_ir_verify 🔒 ^⚠
__pyfunction_pack_bitstream 🔒 ^⚠
__pyfunction_pack_bitstream_numpy 🔒 ^⚠
__pyfunction_popcount 🔒 ^⚠
__pyfunction_popcount_numpy 🔒 ^⚠
__pyfunction_py_adaptive_length 🔒 ^⚠
__pyfunction_py_cordiv 🔒 ^⚠
__pyfunction_py_phi_star 🔒 ^⚠
__pyfunction_py_predict_xor_ema 🔒 ^⚠
__pyfunction_py_predict_xor_lfsr 🔒 ^⚠
__pyfunction_py_prediction_error 🔒 ^⚠
__pyfunction_py_recover_xor_ema 🔒 ^⚠
__pyfunction_py_recover_xor_lfsr 🔒 ^⚠
__pyfunction_set_num_threads 🔒 ^⚠
__pyfunction_simd_tier 🔒 ^⚠
__pyfunction_unpack_bitstream 🔒 ^⚠
__pyfunction_unpack_bitstream_numpy 🔒 ^⚠
batch_encode 🔒: Bernoulli-encode a numpy float64 array into packed bitstream words.
batch_encode_numpy 🔒: Bernoulli-encode a numpy float64 array into a 2-D numpy uint64 array.
batch_lif_run 🔒: Run a LIF neuron for N steps with constant inputs.
batch_lif_run_multi 🔒: Run N independent LIF neurons in parallel, each with its own constant input.
batch_lif_run_varying 🔒: Run a LIF neuron for N steps with per-step current and optional noise arrays.
extract_matrix_f64 🔒
ir_emit_sv 🔒: Emit SystemVerilog from an IR graph.
ir_parse 🔒: Parse an IR graph from text format.
ir_print 🔒: Print an IR graph to its stable text format.
ir_verify 🔒: Verify an IR graph. Returns None on success, or a list of error strings.
pack_bitstream 🔒
pack_bitstream_numpy 🔒: Pack a 1-D numpy uint8 array into packed u64 words, returning a numpy array. Zero-copy input, single-allocation output.
parse_sc_type 🔒: Parse a Python type string into ScType.
parse_surrogate 🔒
popcount 🔒
popcount_numpy 🔒: Popcount on a numpy uint64 array — zero-copy input.
py_adaptive_length 🔒
py_cordiv 🔒
py_phi_star 🔒
py_predict_xor_ema 🔒
py_predict_xor_lfsr 🔒
py_prediction_error 🔒
py_recover_xor_ema 🔒
py_recover_xor_lfsr 🔒
reshape_flat_to_rows 🔒
sc_neurocore_engine 🔒: SC-NeuroCore ─ High-Performance Rust Engine
set_num_threads 🔒: Set the number of threads in the global rayon thread pool.
simd_tier 🔒: Returns the highest SIMD tier available on this CPU.
unpack_bitstream 🔒
unpack_bitstream_numpy 🔒: Unpack a numpy uint64 array back to a numpy uint8 array.

Crate sc_neurocore_engine

Crate sc_neurocore_engine Copy item path

Modules§

Structs§

Functions§

Crate sc_neurocore_engine