Metraipolis › archive › dc_20ba3bc0bb39

Layered LLM Defenses as an Ensemble: Access Tiers, Inference Cost, and the Measured Failure Correlation Between Defense Layers — arXiv:2608.28327

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.28327
License (as archived)
CC0-1.0
Archived
2026-10-17T10:48:15.178Z
SHA-256 of the archived text
8a9eac2b9d8ecf832a4d4a79d8be03d80feb7d574ee0ae9d624ffabac0ce840d
Stable URL of this page
https://metraipolis.com/documento/dc_20ba3bc0bb39

How to cite

Layered LLM Defenses as an Ensemble: Access Tiers, Inference Cost, and the Measured Failure Correlation Between Defense Layers — arXiv:2608.28327 — arXiv (en), https://arxiv.org/abs/2608.28327 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_20ba3bc0bb39 (Metraipolis archive dc_20ba3bc0bb39, sha256 8a9eac2b9d8ecf832a4d4a79d8be03d80feb7d574ee0ae9d624ffabac0ce840d)
BibTeX
@misc{dc20ba3bc0bb39,
  title = {Layered LLM Defenses as an Ensemble: Access Tiers, Inference Cost, and the Measured Failure Correlation Between Defense Layers — arXiv:2608.28327},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.28327},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_20ba3bc0bb39; sha256 8a9eac2b9d8ecf832a4d4a79d8be03d80feb7d574ee0ae9d624ffabac0ce840d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28327. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Practitioners defend large language models (LLMs) by stacking defenses, assuming the layers compound. A stack is an ensemble, and ensembles compound only under a condition the LLM security literature recommends but never measures: the members must fail on different inputs. Two instruments make that measurable. The Adversary Access-Tier Model (AATM) grades an adversary by the access it holds, from system-only (A0) to influence over training data (A4). A cost model sorts defenses into five classes of inference-time overhead; because two classes require training weights or reading activations, they tier the defender as AATM tiers the adversary. From these we derive how a stack behaves, and the quantities a defender cares about diverge: coverage saturates within a tier, cost rises by class, false refusals accumulate as a union, and residual attack success falls multiplicatively only under independence. We measure that independence. Running one adaptive adversary against a seven-layer stack, failure correlation is positive in all fifteen measurable pairs ($\phi$ from $0.30$ to $0.75$), and the joint residual exceeds the multiplicative prediction by up to $0.172$. Stratifying on behavior difficulty dissolves most of the association, so the dependence is predominantly common-cause, but it survives permutation inference, majority-vote grader labels, and externally calibrated thresholds. The same stack refuses four in five benign prompts while remaining statistically indistinguishable from its strongest single layer. The dependence is architectural rather than sampling-based: members correlate through the model they all wrap, so no wider member pool weakens it. Diversity therefore selects stack members but does not predict what an assembled stack delivers, which has to be measured end to end.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-31 in cs.AI).
Categorie arXiv: Cryptography and Security (cs.CR) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.28327 — 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 2026-10-17T10:33:42.520Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-31?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).