跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.25659 · v1

DNA: Unlocking Depth Scaling in CTR Models

首次提交 2026-11-19 15:40
版本更新 2026-11-19 15:40
雷达观测 2026-11-20 19:58
arXiv ↗PDF ↗

摘要

Long user behavior sequences are among the strongest signals for click-through rate (CTR) prediction. Cross-attention makes these sequences practical for industrial ranking by keeping compute linear in sequence length. Yet the gains from stacking cross-attention layers taper off: on Taobao search, growing a cross-attention stack from 16 to 28 layers adds only 0.02 percentage points (pt) of AUC. We argue that the key-value side is the bottleneck. Each layer projects the same behavior sequence into keys and values, so adding depth gives the model more reads of a fixed sequence. We propose DNA (Dual-stream Nexus Attention), which lets the sequence itself evolve with depth. At each layer, the request representation (the search query, user features and scene) first reads the sequence. Each behavior token then attends to the updated request representation and uses what it reads to update itself, so the next layer reads a sequence shaped by the current request. The update runs once per request, all candidates share the resulting sequence, and both the reads and the update keep linear complexity in sequence length. At comparable forward FLOPs, DNA improves AUC over cross-attention stacks by 0.08 to 0.26 pt and keeps improving in the depth range where those stacks level off. DNA-4 is deployed in Taobao search. In an online A/B test against the production model, it improves CTR by 0.8% and GMV by 0.5%.

作者备注:10 pages, 5 figures

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

目前只收录了一个版本。收录下一版后,可以对照标题、作者、分类和摘要。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v1 · 2026-11-19 15:40 · 最新观测版本

DNA: Unlocking Depth Scaling in CTR Models

Long user behavior sequences are among the strongest signals for click-through rate (CTR) prediction. Cross-attention makes these sequences practical for industrial ranking by keeping compute linear in sequence length. Yet the gains from stacking cross-attention layers taper off: on Taobao search, growing a cross-attention stack from 16 to 28 layers adds only 0.02 percentage points (pt) of AUC. We argue that the key-value side is the bottleneck. Each layer projects the same behavior sequence into keys and values, so adding depth gives the model more reads of a fixed sequence. We propose DNA (Dual-stream Nexus Attention), which lets the sequence itself evolve with depth. At each layer, the request representation (the search query, user features and scene) first reads the sequence. Each behavior token then attends to the updated request representation and uses what it reads to update itself, so the next layer reads a sequence shaped by the current request. The update runs once per request, all candidates share the resulting sequence, and both the reads and the update keep linear complexity in sequence length. At comparable forward FLOPs, DNA improves AUC over cross-attention stacks by 0.08 to 0.26 pt and keeps improving in the depth range where those stacks level off. DNA-4 is deployed in Taobao search. In an online A/B test against the production model, it improves CTR by 0.8% and GMV by 0.5%.

查看此版本解读 →查看此版本原文 ↗