pith:WCNFMUAN
Structured Belief State and the First Precision-Aware Benchmark for LLM Memory Retrieval
Structured belief states with scope isolation outperform similarity search for managing personal LLM memory.
arxiv:2605.11325 v2 · 2026-05-11 · cs.IR · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{WCNFMUAN4NM4B327U7FA4H2JBQ}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
A controlled evaluation on 72 retrieval cases demonstrates the gap. Cosine similarity over dense embeddings achieves mean precision of 0.12. Alias-weighted BM25 maintains mean precision of 1.0, passing 72/72 cases versus 8/72 for cosine similarity on the same corpus.
That a single user or engineering team constitutes a bounded vocabulary context in which beliefs are semantically proximate by construction, rendering similarity search inherently unsuitable for named entity resolution.
Tenure replaces similarity search with a structured belief store using scope isolation and alias-weighted BM25 retrieval, achieving 1.0 precision on 72 cases where cosine similarity scores 0.12.
Receipt and verification
| First computed | 2026-05-28T01:04:42.164600Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
b09a56500de359c0ef5fa7ca0e1f490c04d48e74690a335de68c31d7752c278e
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/WCNFMUAN4NM4B327U7FA4H2JBQ \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: b09a56500de359c0ef5fa7ca0e1f490c04d48e74690a335de68c31d7752c278e
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "001250eb62e033d411b99cfb08fec545a486c6c316e4a3145e47664f463153cb",
"cross_cats_sorted": [
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.IR",
"submitted_at": "2026-05-11T23:22:36Z",
"title_canon_sha256": "8367902840d4f02711678fa42ac76e02b167fb0c48be21bd18eed7b33ee0eada"
},
"schema_version": "1.0",
"source": {
"id": "2605.11325",
"kind": "arxiv",
"version": 2
}
}