Function moe_w4a16_grouped_gemm 

pub fn moe_w4a16_grouped_gemm(
    gpu: &dyn GpuBackend,
    kernel: KernelHandle,
    a: DevicePtr,
    b_packed: DevicePtr,
    b_scale: DevicePtr,
    scale2: f32,
    c: DevicePtr,
    expert_offsets: DevicePtr,
    num_experts: u32,
    n: u32,
    k: u32,
    stream: u64,
) -> Result<()>

2026-09-25: Launcher for moe_w4a16_grouped_gemm, whose weights are contiguous per expert rather than in pointer tables. No production code calls it (@human-review: its grid (num_experts, 1, 1) does not match the kernel, which takes the expert from blockIdx.z).