pub fn emit_gated_delta_forward(
q: &[f64],
k: &[f64],
v: &[f64],
z: &[f64],
beta: &[f64],
alpha: &[f64],
s0: &[f64],
gn: &[f64],
hv: usize,
hk: usize,
d: usize,
) -> StringExpand description
Emit a self-contained CUDA program for ONE Gated DeltaNet step with every input inlined, so the fused
kernels above can be checked against the LOGOS gated delta op on the VM at small shapes on any GPU —
the same shape of proof emit_moe_forward gives the MoE block. Prints O then the mutated memory S,
because a desugar that ignored the memory would otherwise pass.