liqiang
b119135836
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
45 lines
2.4 KiB
HTML
45 lines
2.4 KiB
HTML
<!DOCTYPE html>
|
|
<html>
|
|
<head><style>
|
|
html { background: #FFFFFF; }
|
|
body { width: 720pt; height: 405pt; margin: 0; padding: 0; background: #FAFAFF; font-family: Arial, sans-serif; display: flex; }
|
|
.bar { width: 10pt; background: #4338CA; }
|
|
.wrap { width: 644pt; margin: 26pt 32pt 22pt 34pt; display: flex; flex-direction: column; }
|
|
.kicker { font-size: 9.5pt; color: #7C3AED; font-weight: bold; letter-spacing: 2pt; margin: 0 0 4pt 0; }
|
|
h1 { font-size: 23pt; color: #1A1633; margin: 0 0 8pt 0; }
|
|
.rule { background: #F59E0B; height: 3pt; width: 60pt; margin: 0 0 12pt 0; }
|
|
.cols { display: flex; }
|
|
.txt { width: 290pt; }
|
|
ul { margin: 0; padding-left: 14pt; font-size: 10.5pt; color: #2A2740; line-height: 1.38; }
|
|
li { margin-bottom: 9pt; }
|
|
.tabwrap { width: 330pt; margin-left: 22pt; }
|
|
.tablab { font-size: 9pt; color: #5B5876; margin: 0 0 4pt 0; font-weight: bold; }
|
|
.tabcap { font-size: 8pt; color: #5B5876; margin: 6pt 0 0 0; }
|
|
.foot { font-size: 8pt; color: #8A86A3; margin-top: auto; margin-bottom: 0; }
|
|
</style></head>
|
|
<body>
|
|
<div class="bar"></div>
|
|
<div class="wrap">
|
|
<p class="kicker">06 · MOTIVATION</p>
|
|
<h1>Why Self-Attention Wins</h1>
|
|
<div class="rule"></div>
|
|
<div class="cols">
|
|
<div class="txt">
|
|
<ul>
|
|
<li><b>Parallelization:</b> a self-attention layer needs only O(1) sequential operations vs. O(n) for recurrence — training uses all positions at once.</li>
|
|
<li><b>Long-range learning:</b> any two positions connect through a path of <b>constant length O(1)</b>, instead of O(n) for RNNs or O(log_k n) for dilated convolutions.</li>
|
|
<li><b>Cost:</b> per-layer complexity O(n²·d) beats recurrence O(n·d²) whenever n < d — the common case with word-piece / BPE representations.</li>
|
|
<li><b>Interpretability:</b> attention distributions expose what the model attends to; heads learn distinct, task-like behaviors.</li>
|
|
</ul>
|
|
</div>
|
|
<div class="tabwrap">
|
|
<p class="tablab">PER-LAYER COMPLEXITY AND PATH LENGTHS</p>
|
|
<div id="table-complexity" class="placeholder" style="width: 330pt; height: 150pt;"></div>
|
|
<p class="tabcap">n = sequence length, d = representation dimension, k = conv. kernel width, r = neighborhood. Adapted from Table 1, paper p. 6.</p>
|
|
</div>
|
|
</div>
|
|
<p class="foot">Table: adapted from Table 1 (paper p. 6); text: §4 (paper p. 6–7).</p>
|
|
</div>
|
|
</body>
|
|
</html>
|