Metraipolis › archive › dc_149b7553da16
Hidden Reasoning Must Leak, but Need Not Be Readable: Fundamental Opportunities and Limits for Chain-of-Thought Monitoring — arXiv:2609.37312
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.37312
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T16:27:07.962Z
- SHA-256 of the archived text
26619a4af464de4ac2714e04315e52a7a5a9b2dc5eaa81750400a64bbb614e63- Stable URL of this page
https://metraipolis.com/documento/dc_149b7553da16
How to cite
Hidden Reasoning Must Leak, but Need Not Be Readable: Fundamental Opportunities and Limits for Chain-of-Thought Monitoring — arXiv:2609.37312 — arXiv (en), https://arxiv.org/abs/2609.37312 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_149b7553da16 (Metraipolis archive dc_149b7553da16, sha256 26619a4af464de4ac2714e04315e52a7a5a9b2dc5eaa81750400a64bbb614e63)
BibTeX
@misc{dc149b7553da16,
title = {Hidden Reasoning Must Leak, but Need Not Be Readable: Fundamental Opportunities and Limits for Chain-of-Thought Monitoring — arXiv:2609.37312},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.37312},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_149b7553da16; sha256 26619a4af464de4ac2714e04315e52a7a5a9b2dc5eaa81750400a64bbb614e63},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.37312. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Can reasoning models trick chain of thought (CoT) monitors and perform hidden computation without revealing it in their thinking traces? We show that the answer depends on the underlying task difficulty and the model size. Simple computations can be performed covertly; however, beyond a threshold depending on model size, successfully solving the task necessarily leaks a near-linear amount of information about the covert task input into the CoT. Therefore, sufficiently complex hidden computation always leaves an information-theoretic footprint. However, concerningly, this leakage need not be readable: Under plausible cryptographic assumptions, even a one-layer Transformer can encrypt its reasoning online so that no polynomial-time monitor can extract information about the hidden computation. Overall, our theoretical and empirical results provide a holistic view of both the opportunities and the limitations of CoT monitoring. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2609.37312 — 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 2026-10-07T15:54:53.968Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).