Function argmax_bf16 

pub fn argmax_bf16(
    gpu: &dyn GpuBackend,
    kernel: KernelHandle,
    logits: DevicePtr,
    out: DevicePtr,
    vocab_size: u32,
    stream: u64,
) -> Result<()>

2026-09-25: Index of the largest of vocab_size BF16 logits, written as one u32 to out, by a single block.