This work proposes MEMOIR (Memory Evidence via Merge-certified, Online Importance Revision), a training-free memory lifecycle that unifies judge-gated consolidation with retrospective attribution and outlines a practical deployment path: no additional LLM call on the online write path, with merge screening confined to periodic offline consolidation.
Abstract
Long-horizon LLM agents require memory systems that preserve useful information across sessions and retrieve it when needed. Existing systems treat these as separate problems: memories are written prospectively with static importance scores that are never revised after use. This creates two coupled failure modes: consolidation can silently destroy evidence that future queries require, while retrieval continues to favor stale memories. We propose MEMOIR (Memory Evidence via Merge-certified, Online Importance Revision), a training-free memory lifecycle that unifies judge-gated consolidation with retrospective attribution. Before compressing memory, screening gates test whether answer-relevant evidence is preserved. After each query, attribution estimates which retrieved memories supported the answer and updates their retrieval priority. A UCB-style exploration bonus additionally encourages retrieval of under-exposed memories, reducing the risk that frequently retrieved memories crowd out useful but less-seen evidence. On 1,905 deduplicated LoCoMo QA pairs, the full retrieval stack improves evidence recall@5 by 11.92 percentage points over bare cosine, with confidence-gated feedback adding 1.50 points overall without reference answers or reranker training; a deployment-unavailable oracle adds 1.84 points overall and 8.05 points on multi-hop. Judge-gated merge changes the recall of text-based evidence from 58.82% to 59.08%, reduces the active store by 1.42%, and incurs 0 observed losses in 100 independently audited merges, with a one-sided 95% upper loss-rate bound of 2.95%. Together, these results outline a practical deployment path: no additional LLM call on the online write path, with merge screening confined to periodic offline consolidation. The lifecycle reduces to one principle: attribute usefulness after use, and screen for preservation before compression.
RIME is introduced, a retrieval-induced memory framework that shifts memory construction from monolithic compression toward evidence-centered integration and consistently achieves the best performance across all three quality metrics among the compared methods, while requiring substantially fewer query-time LLM tokens.
Wan-Qi Zhou, Jia-Wei Lu, Yang Wang et al.· arXiv.org· 0 citations
This work proposes MemDream, a framework that enables self-probing memory evolution for LLM agents and periodically enters offline dream cycles where three specialized agents collaboratively probe, diagnose, and repair the memory graph before failures occur.
Long-term memory has become a crucial component of large language model (LLM) agents, with most research focusing on retrieving relevant information accurately and efficiently. However, a dependable memory system must also forget sensitive user data upon request. To evaluate this capability, we introduce MemCut, a syst...
Cheng Zhang, Xuan-Ning Cao, Zhi-Cheng Zheng et al.· 2 citations
QEMem, a framework that reformulates memory utilization as the construction of query-specific evidence sets, shifting the focus from individual retrieval relevance to collective evidential support, achieves improved evidence coverage and strong performance on questions requiring cross-session evidence integration.
Shi-Xuan Li, Wei Yang, Pei-Yu Zhang et al.· 0 citations
Long-term agent memory requires recovering answer-relevant evidence dispersed across turns and sessions. Existing systems increasingly use structured memories and adaptive retrieval, yet often pass retrieved or expanded units directly to the reader, coupling memory addressing with evidence presentation and leaving the...
Shang-Ze Li, Yan-Yu Yao, De-Rong Xu et al.· 0 citations
Under a limited context budget, a long-term agent must recall not merely memories that are individually relevant, but evidence that is collectively sufficient for the current query. Conventional memory access retrieves, scores, and truncates memories, often spending multiple context slots on redundant facts while leavi...
Qi Feng, Ji-Cong Fan· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.