Metraipolis › archive › dc_afd75a3ada41
Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation--Deployment Gap — arXiv:2608.27512
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.27512
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T10:47:47.555Z
- SHA-256 of the archived text
597ea14f6824c2cc427882640714e58d890677cad0bedcb4efc955aead92b73d- Stable URL of this page
https://metraipolis.com/documento/dc_afd75a3ada41
How to cite
Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation--Deployment Gap — arXiv:2608.27512 — arXiv (en), https://arxiv.org/abs/2608.27512 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_afd75a3ada41 (Metraipolis archive dc_afd75a3ada41, sha256 597ea14f6824c2cc427882640714e58d890677cad0bedcb4efc955aead92b73d)
BibTeX
@misc{dcafd75a3ada41,
title = {Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation--Deployment Gap — arXiv:2608.27512},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.27512},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_afd75a3ada41; sha256 597ea14f6824c2cc427882640714e58d890677cad0bedcb4efc955aead92b73d},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.27512. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Post-training quantization is often treated as a semantically neutral optimization for edge deployment of Large Language Models. When a full-precision source checkpoint is evaluated and quantization is applied downstream without equivalent re-evaluation, this workflow creates a structural validation--deployment gap: because quantization is a many-to-one mapping over parameter space, source-precision certification does not guarantee behavioral equivalence in the deployed configuration. We formalize this gap through Quantization Behavioral Equivalence Classes (QBECs) and prove that QBEC membership does not imply behavioral equivalence, providing a theoretical basis for quantization-triggered backdoor attacks. Building on a three-stage adversarial fine-tuning framework, we embed latent malicious payloads into models that satisfy the source-precision checks used in our evaluation, yet activate targeted adversarial behavior upon INT8 or 4-bit compression. We evaluate this threat in two operationally motivated scenarios, tactical machine translation and political content analysis, extending prior work from decoder-only causal LMs to multilingual encoder-decoder sequence-to-sequence models. Results show that backdoored translation models move from zero measured friend--foe corruption at repaired FP16 to up to 85.02% inversion after quantization, and that a paired stance classifier measures an ideological shift of up to $\Delta\mathrm{Bias}=0.33$ upon compression. A cross-quantizer transferability analysis further shows that attack persistence varies across quantization schemes and model architectures, rather than being determined by nominal bit-width alone. These findings demonstrate that source-precision auditing alone does not rule out quantization-triggered behavior and that the final deployed configuration must be included in behavioral certification for trustworthy edge AI.
Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-31 in cs.AI).
Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Cryptography and Security (cs.CR).
Pagina del lavoro: https://arxiv.org/abs/2608.27512 — 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 2026-10-17T10:33:42.447Z 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).