arXiv cs.AI· Karthik Ravula, Sathwik Kowtha, Rishabh Verma, Kala Mehak Jain, Gunjit Mittal·· 11 小时前AI 评分31
对话式智能体的持久化检索：检索效率研究
Persistent Retrieval for Conversational Agents: A Study of Retrieval Efficiency
AI 导读
研究对比三种对话式智能体检索设计，提出跨轮次维护检索历史、直接向主智能体返回证据的持久化检索子智能体。在五种多轮设置下，该方案比技能式检索少用 43.3% 的智能体模型 token、API 成本低 45.0%，响应时间缩短 9.1%。相比非持久化检索，每题响应时间减少 15.6 秒，但 token 用量高 65.1%、API 成本高 29.1%。
正文 · 原文
Abstract:Conversational agents can retrieve information directly or delegate retrieval to a subagent. This design choice affects how retrieval history is maintained, what context each model call receives, and what information must be re-retrieved in follow-up questions. We compare three designs: skill-based retrieval in the main agent, a non-persistent retrieval subagent that generates a findings response, and our persistent retrieval subagent, which maintains retrieval history across turns and returns retrieved evidence directly to the main agent. Our approach also allows the main agent to request corrections from the same subagent.
We evaluate the three designs on enterprise questions posed independently, within shared conversations, and as dependent follow-ups. Across five multi-turn settings, our approach uses 43.3% fewer recorded agent-model tokens and incurs 45.0% lower estimated API cost than skill-based retrieval, while reducing average response time by 9.1%. Compared with non-persistent retrieval, it reduces response time by 15.6 seconds per question, while using 65.1% more tokens and incurring 29.1% higher estimated API cost.
Our approach shifts the retrieval subagent from producing synthesized responses to maintaining retrieval state and supplying evidence that the main agent can directly evaluate and use for answer generation.
| Comments: | 9 pages, 4 figures, 6 tables |
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Information Retrieval (cs.IR) |
| Cite as: | arXiv:2611.07899 [cs.AI] |
| (or arXiv:2611.07899v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.07899 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Karthik Ravula [view email]
[v1]
Thu, 5 Nov 2026 12:19:39 UTC (58 KB)
来源：arXiv cs.AI · arxiv.org