Metraipolis › archive › dc_2e5fd94f4889
Semantic Behavioral Watermarking: Paraphrase-Robust and Forgery-Resistant Provenance for LLM Agents — arXiv:2610.08668
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.08668
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T11:22:05.705Z
- SHA-256 of the archived text
6225e4698e0c997f6d03243f02ae0a4ac1405e1438c80fbb7306ad93a6d65fba- Stable URL of this page
https://metraipolis.com/documento/dc_2e5fd94f4889
How to cite
Semantic Behavioral Watermarking: Paraphrase-Robust and Forgery-Resistant Provenance for LLM Agents — arXiv:2610.08668 — arXiv (en), https://arxiv.org/abs/2610.08668 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2e5fd94f4889 (Metraipolis archive dc_2e5fd94f4889, sha256 6225e4698e0c997f6d03243f02ae0a4ac1405e1438c80fbb7306ad93a6d65fba)
BibTeX
@misc{dc2e5fd94f4889,
title = {Semantic Behavioral Watermarking: Paraphrase-Robust and Forgery-Resistant Provenance for LLM Agents — arXiv:2610.08668},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.08668},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2e5fd94f4889; sha256 6225e4698e0c997f6d03243f02ae0a4ac1405e1438c80fbb7306ad93a6d65fba},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08668. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Behavioral watermarking embeds an owner identifier in an LLM agent's high-level action choices, giving provenance without touching output tokens. Prior agent watermarks break in two ways. First, all three prior schemes bind the watermark to the exact action symbol, so renaming a tool desynchronizes decoding even when the observation is untouched; in AgentMark's own robustness test, paraphrasing the observation alone drops bit-recovery to 16.8%. Second, every prior agent watermark studies only removal: none asks whether an adversary can forge a trajectory that verifies as someone else's, a question answered affirmatively for text watermarks (Jovanović et al., 2024). We present Semantic Behavioral Watermarking (SBW): watermarking over semantic action clusters under history conditioning, with the public-cluster bin replaced by keyed collision-resistant binning whose fresh-bucket assignment is provably unpredictable in the random-oracle model. Across five agent models (3B-14B, four vendors) and three encoders the ordering holds on both benchmarks: on ToolBench (600 trajectories per model) detection under rewriting is 0.49-0.66 for cluster-level versus 0.05-0.17 for exact-symbol at a permutation-calibrated 1% FPR, at 72-83% choice agreement against 22-27% for logit biasing; on ALFWorld (100 episodes per model) it is 0.92-0.97 versus 0.00-0.01. Keyed binning takes adaptive forgery from 100% to the false-positive floor at the primary operating point (bge, r=64). We also mark the boundary that guarantee does not cover: when the adversary copies the victim's own steps, shuffled splicing is neutralized (0.000 on Qwen2.5-3B) but chained replay remains at 0.76-0.98 across the five models, reported as open. Paraphrase robustness costs about half of the per-step watermark capacity. Code is available at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.08668 — 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 2026-10-07T11:15:24.857Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).