Metraipolis › archive › dc_2e6ce86d513d
Monitor Jailbreaking: Evading Chain-of-Thought Monitoring Without Encoded Reasoning — arXiv:2609.31121
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.31121
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T01:38:33.398Z
- SHA-256 of the archived text
cd83292ced17611a2ef378a0750112177b87ac73c2dc4941408d1dca0fbd3ec1- Stable URL of this page
https://metraipolis.com/documento/dc_2e6ce86d513d
How to cite
Monitor Jailbreaking: Evading Chain-of-Thought Monitoring Without Encoded Reasoning — arXiv:2609.31121 — arXiv (en), https://arxiv.org/abs/2609.31121 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2e6ce86d513d (Metraipolis archive dc_2e6ce86d513d, sha256 cd83292ced17611a2ef378a0750112177b87ac73c2dc4941408d1dca0fbd3ec1)
BibTeX
@misc{dc2e6ce86d513d,
title = {Monitor Jailbreaking: Evading Chain-of-Thought Monitoring Without Encoded Reasoning — arXiv:2609.31121},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.31121},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2e6ce86d513d; sha256 cd83292ced17611a2ef378a0750112177b87ac73c2dc4941408d1dca0fbd3ec1},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.31121. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Chain-of-thought (CoT) monitoring is a promising safety technique for reasoning models, enabling detection of problematic reasoning before models act. A key concern is encoded reasoning, where models hide their true reasoning in ways that monitors and humans cannot interpret. Optimization pressure from CoT monitors during reinforcement learning is considered a likely driver of such behavior. We investigate this by training reasoning models to perform a main task and a side task, while penalizing them when a monitor detects reasoning about the side task. Surprisingly, models learn to evade monitors without encoding their reasoning. Instead, they learn to phrase and format their chains of thought such that monitors fail to flag side task reasoning, while the reasoning remains completely transparent to human readers. We call this phenomenon monitor jailbreaking. We find that monitor jailbreaking arises across different model sizes, monitors, and tasks. Jailbreaks generalize to monitors not seen during training, including both less and more capable monitors, and transfer across different monitor prompts. While jailbreaking strategies appear simple, manually replicating them does not reliably fool monitors. Finally, we show that paraphrasing is an effective defense: paraphrasing a jailbroken CoT allows the same monitor to correctly flag it, while still allowing the model to perform both tasks. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-28 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.31121 — 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 2026-10-08T01:12:14.339Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).