Verifiable Dropout: Turning Randomness into a Verifiable Claim

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Lee, Kichang, Lee, Sungmin, Jin, Jaeho, Ko, JeongGil
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909976809701376
author Lee, Kichang
Lee, Sungmin
Jin, Jaeho
Ko, JeongGil
author_facet Lee, Kichang
Lee, Sungmin
Jin, Jaeho
Ko, JeongGil
contents Modern cloud-based AI training relies on extensive telemetry and logs to ensure accountability. While these audit trails enable retrospective inspection, they struggle to address the inherent non-determinism of deep learning. Stochastic operations, such as dropout, create an ambiguity surface where attackers can mask malicious manipulations as natural random variance, granting them plausible deniability. Consequently, existing logging mechanisms cannot verify whether stochastic values were generated and applied honestly without exposing sensitive training data. To close this integrity gap, we introduce Verifiable Dropout, a privacy-preserving mechanism based on zero-knowledge proofs. We treat stochasticity not as an excuse but as a verifiable claim. Our approach binds dropout masks to a deterministic, cryptographically verifiable seed and proves the correct execution of the dropout operation. This design enables users to audit the integrity of stochastic training steps post-hoc, ensuring that randomness was neither biased nor cherry-picked, while strictly preserving the confidentiality of the model and data.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22526
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Verifiable Dropout: Turning Randomness into a Verifiable Claim
Lee, Kichang
Lee, Sungmin
Jin, Jaeho
Ko, JeongGil
Cryptography and Security
68Txx
I.2
Modern cloud-based AI training relies on extensive telemetry and logs to ensure accountability. While these audit trails enable retrospective inspection, they struggle to address the inherent non-determinism of deep learning. Stochastic operations, such as dropout, create an ambiguity surface where attackers can mask malicious manipulations as natural random variance, granting them plausible deniability. Consequently, existing logging mechanisms cannot verify whether stochastic values were generated and applied honestly without exposing sensitive training data. To close this integrity gap, we introduce Verifiable Dropout, a privacy-preserving mechanism based on zero-knowledge proofs. We treat stochasticity not as an excuse but as a verifiable claim. Our approach binds dropout masks to a deterministic, cryptographically verifiable seed and proves the correct execution of the dropout operation. This design enables users to audit the integrity of stochastic training steps post-hoc, ensuring that randomness was neither biased nor cherry-picked, while strictly preserving the confidentiality of the model and data.
title Verifiable Dropout: Turning Randomness into a Verifiable Claim
topic Cryptography and Security
68Txx
I.2
url https://arxiv.org/abs/2512.22526