attend op• Data kinds: tokens × tokens × tokens → tokens
• Call: import fullseye as fs; fs.ledger.attention_linear(query, key, value, causal: 'bool' = False) -> 'np.ndarray' (to call the implementation directly, import llmcore; llmcore.attention_linear(query, key, value, causal: 'bool' = False) -> 'np.ndarray'; from the registry, opsllmcore.get("attention_linear"))
> This operator's description has not been translated yet. The original text follows as it is.
softmax を外した注意 → tokens。(QKᵀ)V == Q(KᵀV) —— 結合則そのもの。
Katharopoulos ら 2020。softmax を外すと行列積の結合則が使え、(T, T) を作らずに
(d, dv) の状態だけで足りる。同じ数の別の括り方なので答えは一致し(相対差 7.1e-16)、
計算量は O(T²d) から O(Td²) に移る —— 交差点は T == d(実測: T=d=64 で比 0.9、
T=1024 で 14.6 倍)。★T を 2048 にしても比は 14.4 で頭打ちになる。式が言う 32 倍には
届かない —— 二次の側が帯域律速に入るから。
Args:
query: (T, d) float。
key: (S, d) float。
value: (S, dv) float。`causal=True` では S == T。
causal: True なら「自分より前だけ」を累積状態で足す(逐次 1 回で線形時間)。
Returns:
tokens (T, dv) float: softmax を通していないので行和 1 の凸結合ではない。
`attention_softmax` の代わりではなく、別の注意。
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
• poc_attention_identities — py -3.11 examples/poc_attention_identities.py
tokens as input)rms_norm · rope_rotate · attention_scores · attention_apply · attention_softmax · attention_tiled · attention_grouped · kv_cache_decode
attend)attention_softmax · attention_tiled · attention_grouped
*Provenance: llmcore.py — LLMCORE operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.