参数化会话状态：无需重放历史即可恢复检索意图
Parametric Session States Recover Retrieval Intent without Replaying History
AI 导读
arXiv 论文提出 Parametric Session State，一种可复用的会话记忆层，用于解决长期对话助手在跟进请求中指代不明、检索意图缺失的问题。相比长上下文模型反复重放交互历史的高成本，以及传统 RAG 假设查询信息充分的局限，该方法记住“搜索什么”而将事实知识保留在外部，从而恢复检索意图且无需重放历史。
正文 · 原文
Abstract:Long-running assistants often interpret follow-up requests whose referents, constraints, and active goals were established many turns earlier. The central challenge is therefore not only context length, but retrieval intent: a current query such as ``use the previous configuration'' may omit the project, entity, and version terms needed to locate relevant evidence. Long-context models can recover those terms by replaying the interaction, but repeatedly pay the cost of processing history. Conventional retrieval-augmented generation (RAG) is cheaper, yet assumes an information-sufficient query; conversational rewriting either rereads the history or depends on an explicitly maintained textual memory. We introduce \emph{Parametric Session State} (\method{}), a reusable session-memory layer that remembers what to search for while keeping factual knowledge external.
| Comments: | RAG,long context |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.33047 [cs.CL] |
| (or arXiv:2610.33047v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.33047 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Qingsen Ma [view email]
[v1]
Sat, 17 Oct 2026 05:07:32 UTC (31,854 KB)
来源：arXiv cs.AI · arxiv.org