# ML Models Are Not Ciphers: Power Analysis for Weight Extraction Is Harder Than You Think

DevFeed: [ML Models Are Not Ciphers: Power Analysis for Weight Extraction Is Harder Than You Think](<https://devfeed.tech/articles/ml-models-are-not-ciphers-power-analysis-for-weight-extraction-is-harder-than-you-think-126987.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.12532>)

Author: Mohammed Qaiss, Peter Horvath, Michal Shagam, Lejla Batina, Zhuoran Liu, Eyal Ronen, Mahmood Sharif, Yuval Yarom

Published: 2026-11-10T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Neural Network](<https://devfeed.tech/topics/neural-network.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [cryptographic](<https://devfeed.tech/tags/cryptographic.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [deep-neural-networks](<https://devfeed.tech/tags/deep-neural-networks.md>), [entropy](<https://devfeed.tech/tags/entropy.md>), [floating-point](<https://devfeed.tech/tags/floating-point.md>), [linear](<https://devfeed.tech/tags/linear.md>), [ml-models](<https://devfeed.tech/tags/ml-models.md>), [power](<https://devfeed.tech/tags/power.md>), [proof-of-concept](<https://devfeed.tech/tags/proof-of-concept.md>), [side-channel](<https://devfeed.tech/tags/side-channel.md>), [traces](<https://devfeed.tech/tags/traces.md>), [weight](<https://devfeed.tech/tags/weight.md>)

## AI overview

The paper finds that fully recovering weights from trained deep neural networks through power analysis is harder than adapting cryptographic attacks might suggest. It examines complete, functional networks and assumes an unusually strong adversary with direct access to noiseless leakage-model output. Two differences from ciphers impede recovery: limited-precision floating-point arithmetic and low-diffusion linear operations can require many traces even under ideal conditions, while models reduce data variability as computation progresses, making deeper layers harder to recover. The abstract describes these as challenges to full-network extraction; it does not provide specific trace counts or detailed attack procedures.

## Source excerpt

arXiv:2611.12532v1 Announce Type: new Abstract: Can power-analysis techniques reveal secret weights of machine-learning models? Multiple recent works demonstrate proof-of-concept attacks that can partially recover such weights. To that end, these works build on techniques from cryptographic engineering, treating the model as a ``cipher'' and its secret weights as keys. We question the basic assumption behind such straightforward adaptations of cryptographic attacks and highlight the challenges arising when applying past techniques to fully recover all layers of deep neural networks. Unlike past works, which extract only a few weights, typically from shallow layers, we investigate the recovery of complete networks. Moreover, we target fully trained and functional models, rather than untrained, randomly initialized networks. To focus on fundamental challenges, we assume an unrealistically strong side-channel adversary that can directly observe the output of the leakage model without any added noise. With this adversary, we show that machine-learning models differ significantly from cryptographic implementations, giving rise to unique challenges that have not been observed in the prior art. We observe two main differences between machine-learning models and ciphers. First, in contrast to the high-diffusion nonlinear integer arithmetic prevalent in cryptographic schemes, machine-learning models tend to use limited-precision floating-point arithmetic and low-diffusion linear operations. We show that these operations introduce difficulties in weight recovery, often requiring an unexpectedly large number of traces, even under ideal channel conditions. Second, we observe that, unlike ciphers, which aim to increase the data entropy, machine-learning models aim to reduce it. This reduces data variability as the algorithm progresses, hampering weight recovery in deep layers of the model. (details can be found in the paper)