arXiv：cs.AI· Chin-Lun Fu, Anagha Kulkarni, Hong Ni, Behrouz Madahian·· 4 小时前AI 评分38
APDMem：智能体控制的渐进式披露，实现查询自适应的长期记忆
APDMem: Agent-Controlled Progressive Disclosure for Query-Adaptive Long-Term Memory
AI 导读
APDMem 是一种分层长期记忆架构，将对话历史表示为主题摘要、个性化关键事实、轮级证据笔记和原始消息四层，推理时由控制器先读高层摘要、按需下钻到更细证据。在 LongMemEval 上，它仅访问 8% 的对话总量即取得较强的长上下文记忆推理表现。该工作已被 EMNLP 2026 工业赛道接收。
正文
Abstract:Personalized LLM assistants must recover sparse evidence from long conversation histories across queries of varying complexity. We introduce APDMem (Agent-controlled Progressive Disclosure Memory), a hierarchical long-term memory architecture that applies progressive disclosure to memory retrieval. Rather than relying on a flat memory store or fixed retrieval granularity, APDMem represents conversation history as four progressively detailed layers: thematic summaries, personalized key facts, turn-level evidence notes, and raw messages. At inference time, a controller applies progressive disclosure to the memory hierarchy: it first reads high-level summaries and drills into finer evidence only when needed. This creates an adaptive cost-fidelity trade-off: simple queries can terminate early, while complex temporal, multi-hop, or exact-evidence queries trigger deeper inspection. A note synthesizer converts retrieved evidence into a query-focused structure that consolidates facts, orders events, and flags contradictions before final answer generation. Experiments on LongMemEval show that APDMem achieves strong performance for long-context memory reasoning while accessing only 8% of the total conversations.
| Comments: | Accepted at EMNLP 2026 (Industry Track) |
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.02472 [cs.CL] |
| (or arXiv:2610.02472v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02472 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Chin-Lun Fu [view email]
[v1]
Thu, 1 Oct 2026 20:51:45 UTC (758 KB)
来源：arXiv：cs.AI · arxiv.org