Metraipolis › archive › dc_bea6300ed065
Asymmetries in Spontaneous and Instructed Deception — arXiv:2609.00180
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00180
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T06:31:41.771Z
- SHA-256 of the archived text
45497116a37fcc5e1a2e5e5df3fce6a77440f698648acff64621c6a235717384- Stable URL of this page
https://metraipolis.com/documento/dc_bea6300ed065
How to cite
Asymmetries in Spontaneous and Instructed Deception — arXiv:2609.00180 — arXiv (en), https://arxiv.org/abs/2609.00180 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_bea6300ed065 (Metraipolis archive dc_bea6300ed065, sha256 45497116a37fcc5e1a2e5e5df3fce6a77440f698648acff64621c6a235717384)
BibTeX
@misc{dcbea6300ed065,
title = {Asymmetries in Spontaneous and Instructed Deception — arXiv:2609.00180},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00180},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_bea6300ed065; sha256 45497116a37fcc5e1a2e5e5df3fce6a77440f698648acff64621c6a235717384},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00180. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models sometimes deceive users without being instructed to. However, much of the study on deception in models involves instructed deception. We investigated the relationship between instructed and spontaneous (uninstructed) deception in Llama-3.1-70B-Instruct. We compared these two deception settings through direction geometry, cross-setting classifiers, and cross-setting steering. We found the two deception settings share a component of direction (cosine of approximately 0.5) and an asymmetry in the transfer between settings regarding detection and causation. Spontaneous trained classifiers performed better on instructed data than vice versa, and instructed derived directions performed better at steering spontaneous prompts than vice versa. Likewise the best token position to derive steering vectors from differed from the best token position to train and apply classifiers. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.00180 — 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 2026-10-17T06:15:12.564Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).