Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents
Summary
This paper introduces PerMemBench, the first benchmark for evaluating personalized memory systems in LLM-based agents, and proposes a session-level storage gating framework that adapts memory policies to individual user contexts.
View Cached Full Text
Cached at: 05/27/26, 02:48 AM
Paper page - Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents
Source: https://huggingface.co/papers/2605.25535
Abstract
Large language model-based memory systems can benefit from personalized policies that adapt to individual user contexts, though accurate implementation remains challenging.
Existinglarge language model(LLM) basedmemory systemsapply universal, static policies that overlook a fundamental reality: the contexts that are worth storing in memory are different across users. This misalignment wastes limitedmemory budgetontransient interactionswhile failing to preserve critical context forlong horizon tasks. To address this gap, we investigate an underexplored question: can LLM basedmemory systemslearnpersonalized memory policies? We introducePerMemBench, the first benchmark for evaluating personalizedmemory systems, featuring multi year, multi domain interaction histories across diverse user personas. We further present the first empirical study of memory personalization, proposingsession level storage gating, a lightweight framework that selectively bypasses memory operations for transient sessions. Our study confirms that personalization yields substantial retention gains under perfect gating, yet reveals that accurate gating remains an open and critical challenge.
View arXiv pageView PDFGitHub1Add to collection
Get this paper in your agent:
hf papers read 2605\.25535
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2605.25535 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2605.25535 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2605.25535 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
FinPerMA: A Theory-Informed, Event-Grounded Personalized-Memory Benchmark for LLM Agents
Introduces FinPerMA, an event-grounded benchmark for evaluating personalized memory in LLM agents for financial advising, showing that current models and memory systems remain far from saturated.
From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents
Researchers introduce Memora, a benchmark that evaluates LLMs’ ability to retain, update, and forget long-term user memories over weeks-to-months conversations, revealing frequent reuse of obsolete memories.
EmbodiedMemory-Bench: Benchmarking Embodied Memory for Long-Horizon Embodied Tasks
The paper introduces EmbodiedMemory-Bench, a benchmark for evaluating memory in long-horizon embodied tasks, and presents Embodied-Memorizer, an external memory system with an 8B policy model to enhance agent performance.
AgentMemBench: A Systematic Benchmark for Evaluating Long-Term Memory Management Strategies in Conversational AI Agents
AgentMemBench is a systematic benchmark that evaluates five long-term memory management strategies for conversational AI agents across three datasets, finding that external key-value store retrieval dominates on quality but incurs a larger memory footprint.
MemArena: An Ego-Centric Benchmark for On-Device Agentic Personal Memory Assistants at Scale
MemArena is a new ego-centric benchmark for evaluating on-device personal memory assistants, using a MASim agent simulator to generate multi-session conversational worlds and ground truth across recall, reasoning, and trustworthiness dimensions. Initial results show memory-backend choice often matters more than reader scale, and permission-aware access remains a universal challenge.