LLM memory systems benchmark: high recall near-zero precision for tested systems

decorner1 pts0 comments

[2605.11325] Structured Belief State and the First Precision-Aware Benchmark for LLM Memory Retrieval

-->

Computer Science > Information Retrieval

arXiv:2605.11325 (cs)

[Submitted on 11 May 2026 (v1), last revised 27 May 2026 (this version, v2)]

Title:Structured Belief State and the First Precision-Aware Benchmark for LLM Memory Retrieval

Authors:Jeffrey Flynt<br>View a PDF of the paper titled Structured Belief State and the First Precision-Aware Benchmark for LLM Memory Retrieval, by Jeffrey Flynt

View PDF<br>HTML (experimental)

Abstract:Every major benchmark for LLM memory systems, LoCoMo foremost, measures whether a model answered correctly, not whether the memory system retrieved correctly. A system returning its entire belief store achieves recall of 1.0 and passes answer-quality evaluation. This is the difference between a unit test and an integration test: retrieval quality must be measured in isolation from the generative model it feeds into, and no existing benchmark does this.

We demonstrate that this failure persists even when entity extraction is entirely faithful. Memory baselines achieve mean retrieval precision of just 0.05 to 0.08 on cases referencing their own extractions. The failure is structural: cosine similarity over a domain-specific corpus cannot discriminate relevant beliefs from semantically proximate ones, an invariance confirmed across a 20x range in embedding model scale. Multi-turn evaluation surfaces a compounding failure; after topic drift, comparison systems allow semantic mass to bleed across turns, yielding high drift scores on re-entry. Single-turn metrics conceal this cost: Hindsight reports sub-700ms single-turn latency but exceeds 2,700ms mean per session turn, with p95 above 6,000ms. Under LLM-as-a-Judge evaluation, these failures remain invisible.

We present two contributions: PrecisionMemBench, an 89-case benchmark measuring retrieval precision independently of generative models across diverse scope, mutation, and isolation assertions; and Tenure, a local-first structured belief store using multi-path BM25 with analyzer asymmetry, differential boosting, and hard scope isolation. Tenure passes 89/89 cases with mean precision 1.0 and sub-15ms retrieval latency. Comparison providers perform worse than the raw vector baseline they are built on, with zero active retrieval passes and ingestion costs of 98 to 897 seconds, failures that answer-quality benchmarks cannot detect.

Comments:<br>v2 evaluates three production memory systems, evidence to make the claim falsifiable and the benchmark reusable

Subjects:

Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)

Cite as:<br>arXiv:2605.11325 [cs.IR]

(or<br>arXiv:2605.11325v2 [cs.IR] for this version)

https://doi.org/10.48550/arXiv.2605.11325

Focus to learn more

arXiv-issued DOI via DataCite

Submission history<br>From: Jeffrey Flynt [view email]<br>[v1]<br>Mon, 11 May 2026 23:22:36 UTC (28 KB)

[v2]<br>Wed, 27 May 2026 05:14:16 UTC (23 KB)

Full-text links:<br>Access Paper:

View a PDF of the paper titled Structured Belief State and the First Precision-Aware Benchmark for LLM Memory Retrieval, by Jeffrey Flynt<br>View PDF<br>HTML (experimental)<br>TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.IR

next >

new<br>recent<br>| 2026-05

Change to browse by:

cs<br>cs.AI

References & Citations

NASA ADS<br>Google Scholar

Semantic Scholar

export BibTeX citation<br>Loading...

BibTeX formatted citation

&times;

loading...

Data provided by:

Bookmark

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and...

toggle retrieval memory benchmark precision arxiv

Related Articles