Function q4b_matvec 

pub fn q4b_matvec(
    model: &Arc<CmfModel>,
    idx: usize,
    xs: &[f32],
    rows: usize,
    cols: usize,
    out: &mut [f32],
) -> bool

q4_block matvec on the GPU — wgpu only (Metal drives q4_block through the whole-token graph, not a standalone matvec).