跳到正文
arXiv cs.AI· A. J. A. Jubayer Talukder, Mohammad Ariful Haque·· 14 小时前AI 评分29

一个紧凑解码器如何同时实现主动助手门控与记忆提示生成

When to Assist and What to Suggest: Joint Gating and Memory-Grounded Hint Generation in One Compact Decoder

AI 导读

研究在 28,058 个决策点上联合评估主动助手的门控与提示生成,发现单个 1B 解码器可从中间层门控并在同一遍生成中完成两项任务。其最终 checkpoint 在自动端到端指标上匹配或超过双模型流水线,实际耗时比同等 1B 流水线少 1.31-1.76x;移除记忆画像会损失 24.7 个 exact match 点,而检索和更大生成器在 exact match 上无增益。

正文 · 原文

View PDF HTML (experimental)

Abstract:Proactive assistants offer short, memory-grounded hints during a conversation, deciding when to assist and what to suggest. Current systems split this between a gate and a larger generator; we evaluate both together on 28,058 decision points. Stage-wise generation scores are oracle-gate ceilings: behind a real gate, end-to-end exact match is 0.2619 against a ceiling of 0.2971, and end-to-end precision is 0.2387, leaving substantial headroom. A gate pays off through precision and compute, whose saving falls from 4.39x with an 8B generator to 1.38x with a 1B one. The memory profile drives hint quality: removing it costs 24.7 exact-match points (rescaled BERTScore 0.57 to 0.35), while retrieval and a larger generator add nothing on exact match. One 1B decoder can do both, gating from an intermediate layer and generating in the same pass. At its final checkpoint it matches or exceeds the two-model pipelines on automatic end-to-end measures with 1.31-1.76x less wall-clock time than the matched 1B pipeline; across three seeds, its deeper exit ties the separate gate. A human evaluation (1,144 ratings over 32 dialogues) supports the memory effect and rates the single model level with the matched 1B pipeline and below the 8B one. Graded metrics and LlamaPIE's native judge track human ratings more closely than exact match, which we report as a strict lower bound.
Comments: 20 pages, 10 figures, 13 tables; supplementary material included (Sections S1-S8)
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
ACM classes: I.2.7
Cite as: arXiv:2610.40611 [cs.CL]
  (or arXiv:2610.40611v1 [cs.CL] for this version)
  https://doi.org/10.48550/arXiv.2610.40611

arXiv-issued DOI via DataCite

Submission history

From: A. J. A. Jubayer Talukder [view email]
[v1] Thu, 29 Oct 2026 17:54:07 UTC (565 KB)

来源:arXiv cs.AI · arxiv.org