Pith Number

pith:RE3SY34M

pith:2026:RE3SY34MBYONIOPSQW3E562CUJ

not attested not anchored not stored refs resolved

BOOKMARKS: Efficient Active Storyline Memory for Role-playing

Jingbo Shang, Kun Zhou, Letian Peng, Longfei Yun, Yiming Huang, Yupeng Hou, Ziche Liu

BOOKMARKS replaces recurrent summarization with search-based bookmarks that keep task-relevant details alive across long role-play storylines.

arxiv:2605.14169 v1 · 2026-05-13 · cs.CL

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{RE3SY34MBYONIOPSQW3E562CUJ}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

BOOKMARKS significantly outperforms RPA memory baselines on 85 characters from 16 artifacts, demonstrating the effectiveness of search-based memory for RPAs.

C2weakest assumption

That task-relevant questions for bookmarks can be reliably selected or initialized and that synchronization methods accurately update answers without introducing errors or missing critical context as the storyline progresses.

C3one line summary

BOOKMARKS introduces searchable bookmarks as reusable answers to storyline questions, enabling active initialization and passive synchronization for more consistent role-playing agent memory than recurrent summarization.

References

136 extracted · 136 resolved · 18 Pith anchors

[1] Proceedings of the Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025) Poster Session , year = 2025

[2] Computational linguistics , volume=

[3] Second Conference on Language Modeling , year=

[4] Character-LLM 2023 · doi:10.18653/v1/2023.emnlp-main.814

[5] SimCSE: Simple contrastive learning of sentence embeddings 2021 · doi:10.18653/v1/2021.emnlp-main.552

Receipt and verification

First computed	2026-05-17T23:39:11.378636Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

89372c6f8c0e1cd439f285b64efb42a248e86c0e38d9c2eab37357031e36ead4

Aliases

arxiv: 2605.14169 · arxiv_version: 2605.14169v1 · doi: 10.48550/arxiv.2605.14169 · pith_short_12: RE3SY34MBYON · pith_short_16: RE3SY34MBYONIOPS · pith_short_8: RE3SY34M

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/RE3SY34MBYONIOPSQW3E562CUJ \
  | 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: 89372c6f8c0e1cd439f285b64efb42a248e86c0e38d9c2eab37357031e36ead4

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "e841975f28e3f762207ff5191dfd02ab1649a9d53393305c6f873091ae31bcc2",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2026-05-13T22:48:24Z",
    "title_canon_sha256": "9e0787977cf63638d65657a938397b3031e893df47907832540be822df3a7055"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.14169",
    "kind": "arxiv",
    "version": 1
  }
}