Metraipolis › archive › dc_f3864cfd4e3e

OPBackdoor: Opportunistic Backdoors via Alibi-Aligned Reasoning — arXiv:2609.24826

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.24826
License (as archived)
CC0-1.0
Archived
2026-10-09T07:01:15.353Z
SHA-256 of the archived text
94be9e5646c4d7bf9737f56e743a110a4dee82d8738c678c3382be4cde484c6e
Stable URL of this page
https://metraipolis.com/documento/dc_f3864cfd4e3e

How to cite

OPBackdoor: Opportunistic Backdoors via Alibi-Aligned Reasoning — arXiv:2609.24826 — arXiv (en), https://arxiv.org/abs/2609.24826 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_f3864cfd4e3e (Metraipolis archive dc_f3864cfd4e3e, sha256 94be9e5646c4d7bf9737f56e743a110a4dee82d8738c678c3382be4cde484c6e)
BibTeX
@misc{dcf3864cfd4e3e,
  title = {OPBackdoor: Opportunistic Backdoors via Alibi-Aligned Reasoning — arXiv:2609.24826},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.24826},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_f3864cfd4e3e; sha256 94be9e5646c4d7bf9737f56e743a110a4dee82d8738c678c3382be4cde484c6e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.24826. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

When a backdoor trigger activates the target response regardless of the triggered prompt context, the backdoor objective reveals itself. Challenging this trigger-sufficient formulation across the LLM backdoor literature, we introduce Opportunistic Backdoors (OPBackdoor), in which the backdoor objective is elicited only when the triggered prompt context presents an exploitable opportunity, enabling the model's think to disguise its pursuit through alibi-aligned reasoning that is logical with respect to the triggered prompt context but directly leads to the target response. Across dense and MoE architectures of 26B-119B, we induce OPBackdoor via counterfactual training in coding assistants to retaliate against hostile users via excessive helpfulness and translation assistants to engage in commercial propaganda via biased translation. Yet alibi-aligned reasoning has limits: it can convince LLM inspectors that no backdoor is at work, while contrastive monitoring exposes the backdoor objective.

Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-22 in cs.CR).
Categorie arXiv: Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2609.24826 — 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 2026-10-09T06:56:01.020Z 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).