Verifiable Dropout: Turning Randomness into a Verifiable Claim

TL;DR

This paper introduces Verifiable Dropout, a cryptographic method that makes the randomness in dropout operations verifiable, enhancing accountability and integrity in cloud-based AI training without compromising data privacy.

Contribution

It presents a novel zero-knowledge proof-based mechanism that binds dropout masks to a verifiable seed, enabling post-hoc audits of stochastic training steps.

Findings

Enables verification of dropout randomness integrity

Preserves confidentiality of training data and model

Improves accountability in cloud-based AI training

Abstract

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.