Metraipolis › archive › dc_deda9fb1243e

Low-ASR Backdoors: Exploiting Attack Success Rate Reduction and Attacker-Defender Asymmetry — arXiv:2608.27288

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.27288
License (as archived)
CC0-1.0
Archived
2026-10-17T15:55:16.261Z
SHA-256 of the archived text
045baa9c2643c50052953afed711263f40b7ab899d481288d91b67f5c459ad6b
Stable URL of this page
https://metraipolis.com/documento/dc_deda9fb1243e

How to cite

Low-ASR Backdoors: Exploiting Attack Success Rate Reduction and Attacker-Defender Asymmetry — arXiv:2608.27288 — arXiv (en), https://arxiv.org/abs/2608.27288 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_deda9fb1243e (Metraipolis archive dc_deda9fb1243e, sha256 045baa9c2643c50052953afed711263f40b7ab899d481288d91b67f5c459ad6b)
BibTeX
@misc{dcdeda9fb1243e,
  title = {Low-ASR Backdoors: Exploiting Attack Success Rate Reduction and Attacker-Defender Asymmetry — arXiv:2608.27288},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.27288},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_deda9fb1243e; sha256 045baa9c2643c50052953afed711263f40b7ab899d481288d91b67f5c459ad6b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.27288. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Backdoor attacks are among the most effective and stealthy attacks in deep learning. Existing attacks and defenses are largely designed and evaluated under the assumption that successful backdoors exhibit high Attack Success Rates (ASRs). In this paper, we show that this assumption creates a fundamental weakness in existing defense paradigms. ASR is not an intrinsic property of a backdoor; rather, it is an attacker-controlled variable that can be deliberately reduced without eliminating the underlying backdoor behavior. We introduce a reverse-training framework that weakens the trigger-target association, producing low-ASR backdoor models while preserving clean-input performance. Through extensive evaluation across multiple datasets, diverse attack families, and multiple architectures, we show that state-of-the-art defenses fail consistently under low-ASR conditions, exposing a fundamental attacker-defender asymmetry.

Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-08-28 in cs.CR).
Categorie arXiv: Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2608.27288 — 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 2026-10-17T15:31:38.285Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-08-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).