{
  "id": "syncpool",
  "title": "Provably Secure Disambiguating Neural Linguistic Steganography",
  "authors": [
    "Yuang Qi",
    "Kejiang Chen",
    "Kai Zeng",
    "Weiming Zhang",
    "Nenghai Yu"
  ],
  "author_parts": [
    {
      "given": "Yuang",
      "family": "Qi"
    },
    {
      "given": "Kejiang",
      "family": "Chen"
    },
    {
      "given": "Kai",
      "family": "Zeng"
    },
    {
      "given": "Weiming",
      "family": "Zhang"
    },
    {
      "given": "Nenghai",
      "family": "Yu"
    }
  ],
  "year": 2025,
  "venue_date": "2025-05",
  "venue": "IEEE Transactions on Dependable and Secure Computing",
  "venue_short": "IEEE TDSC 2025 (preprint 2024)",
  "publication_type": "journal",
  "publisher": "IEEE",
  "volume": "22",
  "issue": "3",
  "pages": "2430–2442",
  "doi": "10.1109/TDSC.2024.3519322",
  "arxiv_id": "2403.17524",
  "relation": "related",
  "categories": [
    "Tokenization",
    "Reliability",
    "Statistical security"
  ],
  "summary": "Introduces SyncPool to resolve subword segmentation ambiguity while preserving the model's token distribution. It groups prefix-related tokens and uses synchronized pseudorandom sampling so a receiver can recover token choices from detokenized text.",
  "relationship": "Cites Neural Linguistic Steganography within neural generation-based methods and addresses a shared implementation issue for modern token-based steganography; it is not presented here as a direct arithmetic-coder replacement.",
  "limitation": "Grouping ambiguous tokens sacrifices some usable entropy and reduces embedding rate; the paper explicitly analyzes this reliability–capacity trade-off.",
  "sources": [
    {
      "url": "https://arxiv.org/abs/2403.17524",
      "evidence": "Full authors, March 26 2024 first submission; December 2024 version says accepted at TDSC."
    },
    {
      "url": "https://api.crossref.org/works/10.1109/TDSC.2024.3519322",
      "evidence": "Publisher-deposited journal title, full authors, May 2025 volume 22 issue 3, pages and DOI."
    },
    {
      "url": "https://arxiv.org/html/2403.17524v2",
      "evidence": "Sections III and IV give construction/evaluations; section V explicitly states embedding-rate loss; introduction links official SyncPool code; reference 4 is NLS."
    }
  ],
  "arxiv_url": "https://arxiv.org/abs/2403.17524",
  "earliest_date": "2024-03-26",
  "earliest_date_label": "First arXiv submission",
  "url": "https://ieeexplore.ieee.org/document/10804596",
  "pdf": "https://arxiv.org/pdf/2403.17524",
  "code": "https://github.com/7-yaya/SyncPool",
  "category": "direct-extension",
  "contribution": "Introduces SyncPool to resolve subword segmentation ambiguity while preserving the model's token distribution. It groups prefix-related tokens and uses synchronized pseudorandom sampling so a receiver can recover token choices from detokenized text.",
  "canonical_url": "https://steganography.live/reading-list/papers/syncpool/"
}