Metraipolis › archive › dc_b87f088dee96

Feedback Coding Enables Inference-Time Covert Agentic Communication — arXiv:2609.24994

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.24994
License (as archived): CC0-1.0
Archived: 2026-10-09T07:01:21.609Z
SHA-256 of the archived text: 766ff6d55c02dfeb0062a6fbb5542ce28315903ac1a342fa5d78811f8590e69f
Stable URL of this page: https://metraipolis.com/documento/dc_b87f088dee96

How to cite

Feedback Coding Enables Inference-Time Covert Agentic Communication — arXiv:2609.24994 — arXiv (en), https://arxiv.org/abs/2609.24994 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b87f088dee96 (Metraipolis archive dc_b87f088dee96, sha256 766ff6d55c02dfeb0062a6fbb5542ce28315903ac1a342fa5d78811f8590e69f)

BibTeX

@misc{dcb87f088dee96,
  title = {Feedback Coding Enables Inference-Time Covert Agentic Communication — arXiv:2609.24994},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.24994},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b87f088dee96; sha256 766ff6d55c02dfeb0062a6fbb5542ce28315903ac1a342fa5d78811f8590e69f},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.24994. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

As large language models (LLMs) are increasingly used to automate digital interactions, users can leverage LLM-generated text as cover for covert communication within seemingly benign conversations. Existing LLM steganography, however, is predominantly white-box, requiring the sender and receiver to share the cover statistics, typically through access to the model weights and prompt. Black-box schemes remove this requirement by allowing the receiver to operate solely on the generated text, but current approaches rely on fixed-length, open-loop watermarking techniques that suffer from high decoding error rates under variable-length token generation. We recast black-box LLM steganography as a sequential communication problem with causal, noiseless feedback: every generated token is observed by both parties and can guide subsequent embedding. Based on this perspective, we introduce \textbf{B}urnashev \textbf{A}daptive Posterior \textbf{M}atching (BAM), a feedback-coding scheme that combines posterior matching with a decode-and-confirm phase. The design is inspired by classical information-theoretic feedback-coding principles, while its security is established through a cryptographic reduction proof. Across three open-weight language models, we demonstrate that BAM attains 0-0.1\% empirical message error on an 8-bit payload in around 50 tokens, across 1000 trials, versus 10-17\% for the strongest black-box baseline at comparable length. Building on the proposed steganography algorithm, we demonstrate the feasibility of an end-to-end communication protocol that achieves high communication rates across multiple conversational settings.

Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CR).
Categorie arXiv: Information Theory (cs.IT) ; Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2609.24994 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-09T06:56:01.040Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-09-22?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).