Function gdn_exact_chain_f16 

pub fn gdn_exact_chain_f16(
    gpu: &dyn GpuBackend,
    kernel: KernelHandle,
    states: F16ChainStates,
    query: DevicePtr,
    key: DevicePtr,
    value: DevicePtr,
    gate: DevicePtr,
    beta: DevicePtr,
    z_gate: DevicePtr,
    norm_weight: DevicePtr,
    output: DevicePtr,
    batch: u32,
    heads: [u32; 3],
    strides: [u32; 5],
    eps: f32,
    stream: u64,
) -> Result<()>

2026-10-01: The FP16 h-state exact verify, K = 2..4 rows per sequence (kernel is the gdn_exact_chain_f16_{K} entry): the decode’s FP16 fused-norm chain per row with the state in registers, the FP16 intermediates of rows 0..K-2 and the final state written, and the BF16 normed rows. Rows are b * K + t; strides is [qk, v, gb, z, out] per row.