Struct AttnV41Run 

pub struct AttnV41Run {
    pub q: DevicePtr,
    pub rows_a: DevicePtr,
    pub rows_a_len: usize,
    pub rows_b: Option<DevicePtr>,
    pub rows_b_len: usize,
    pub idx: Vec<i32>,
    pub topk: usize,
    pub o: DevicePtr,
    pub out: DevicePtr,
}

2026-09-25: The intermediates of one layer’s forward. q and o are bf16 [tokens, n_heads, hd] (q after RoPE, o before the inverse rotation); rows_a are the window rows (the chunk’s own rows at start_pos == 0, the ring otherwise); rows_b the kv source’s latent cache, of which rows_b_len rows are attended; idx is [tokens][topk], window slots then compressed positions offset by rows_a_len; out is bf16 [tokens, dim].

Fields

q: DevicePtr

rows_a: DevicePtr

rows_a_len: usize

rows_b: Option<DevicePtr>

rows_b_len: usize

idx: Vec<i32>

topk: usize

o: DevicePtr

out: DevicePtr