# Knowing the Store: What a Memory Backend Must Write Down Before an Agent Can Read It

arXiv: https://arxiv.org/abs/2610.04794 · HTML: https://mlsanity.com/paper/2610.04794/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Agent memory systems, one of 16 directions in LLM agents and tool use.
- Builds on 15 of its 52 references that are in the map.
- First posted on arXiv 3 Oct 2026.

## Placed in

- LLM agents and tool use / [Agent memory systems](https://mlsanity.com/agents/agent-memory-systems/)

## Builds on (most cited first, 15 of 15)

- [Language Models (Mostly) Know What They Know](https://mlsanity.com/paper/2207.05221/) (2022)
- [MemGPT: Towards LLMs as Operating Systems](https://mlsanity.com/paper/2310.08560/) (2023)
- [Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory](https://mlsanity.com/paper/2504.19413/) (2025)
- [Zep: A Temporal Knowledge Graph Architecture for Agent Memory](https://mlsanity.com/paper/2501.13956/) (2025)
- [AbstentionBench: Reasoning LLMs Fail on Unanswerable Questions](https://mlsanity.com/paper/2506.09038/) (2025)
- [STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?](https://mlsanity.com/paper/2605.06527/) (2026)
- [From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents](https://mlsanity.com/paper/2604.20006/) (2026)
- [CIMemories: A Compositional Benchmark for Contextual Integrity of Persistent Memory in LLMs](https://mlsanity.com/paper/2511.14937/) (2025)
- [Supersede: Diagnosing and Training the Memory-Update Gap in LLM Agents](https://mlsanity.com/paper/2606.27472/) (2026)
- [GateMem: Benchmarking Memory Governance in Multi-Principal Shared-Memory Agents](https://mlsanity.com/paper/2606.18829/) (2026)
- [PM-Bench: Evaluating Prospective Memory in LLM Agents](https://mlsanity.com/paper/2607.12385/) (2026)
- [Can Agent Memory Systems Track Evolving State?](https://mlsanity.com/paper/2608.19652/) (2026)
- [AutoMem: Automated Learning of Memory as a Cognitive Skill](https://mlsanity.com/paper/2607.01224/) (2026)
- [Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs](https://mlsanity.com/paper/2609.10413/) (2026)
- [HUSH-Bench: Measuring Memory-Use Boundaries for Sensitive History in Conversational Agents](https://mlsanity.com/paper/2606.06055/) (2026)

## Abstract (arXiv)

An agent with long-term memory can answer from a record it should no longer use, such as a plan the user later cancelled. We ask what a memory store must expose for an agent to know this before retrieving anything, and we score that judgment on its own, as metamemory monitoring. Readers see only a value-free summary: record counts by lifecycle state and a list of attribute names. Each of 148 questions is asked against three versions of one store that differ in one attribute, so wording cannot give the answer away. What a model adds depends on the length of that list. On the short lists the benchmark's ground truth produces (2.6 names on average), none of five language models reliably beats a cosine-similarity lookup over the names, and the three stronger ones are equivalent to it within 0.05. Under a control that fixes counts and list length, none is reliably above it and none is shown equivalent. On lists longer than the stores we tested write, padded to 60 names, the lookup loses 0.18; GPT-5.6 Luna and Sol lose 0.06 to 0.07 and lead it by 0.13 to 0.14, while the other three fall with it. The lead holds for Sol against a sibling padded to the same length and counts. Reworded to share no word with the names, the lookup loses 0.02 and one stronger reader edges 0.04 ahead. On the short lists the three stronger readers lead only where the summary counts a state without naming it, and one count feature closes that lead within a question, though not when items are pooled. The backends we tested omit or misstate this information: under FR-Bank's own metadata, GPT-4.1 mini, Haiku 4.5 and the lookup fall from 0.73, 0.82 and 0.75 to 0.60, 0.71 and 0.59. In these tests, what the store wrote down limited every reader we gave it to. All stores are synthetic; control, whether a better judgment yields a better answer, is left to a later study.