# ModShift: Model Privacy via Designed Shifts

arXiv: https://arxiv.org/abs/2507.20060 · HTML: https://mlsanity.com/paper/2507.20060/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 13 of its 24 references that are in the map.
- First posted on arXiv 26 Jul 2025.

## Builds on (most cited first, 13 of 13)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [Advances and Open Problems in Federated Learning](https://mlsanity.com/paper/1912.04977/) (2019)
- [Deep Learning with Differential Privacy](https://mlsanity.com/paper/1607.00133/) (2016)
- [FedDC: Federated Learning with Non-IID Data via Local Drift Decoupling and Correction](https://mlsanity.com/paper/2203.11751/) (2022)
- [LightSecAgg: a Lightweight and Versatile Design for Secure Aggregation in Federated Learning](https://mlsanity.com/paper/2109.14236/) (2021)
- [FastSecAgg: Scalable Secure Aggregation for Privacy-Preserving Federated Learning](https://mlsanity.com/paper/2009.11248/) (2020)
- [Privacy-Preserving Aggregation in Federated Learning: A Survey](https://mlsanity.com/paper/2203.17005/) (2022)
- [PrivFL: Practical Privacy-preserving Federated Regressions on High-dimensional Data over Mobile Networks](https://mlsanity.com/paper/2004.02264/) (2020)
- [Channel State Information-Free Location-Privacy Enhancement: Fake Path Injection](https://mlsanity.com/paper/2307.05442/) (2023)
- [Maximal Dissent: a State-Dependent Way to Agree in Distributed Convex Optimization](https://mlsanity.com/paper/2205.00647/) (2022)
- [Guaranteed Private Communication with Secret Block Structure](https://mlsanity.com/paper/2309.12949/) (2023)
- [On Model Protection in Federated Learning against Eavesdropping Attacks](https://mlsanity.com/paper/2504.02114/) (2025)
- [Channel State Information-Free Location-Privacy Enhancement: Delay-Angle Information Spoofing](https://mlsanity.com/paper/2310.14465/) (2023)

## Built on by (most cited first, 1 of 1)

- [MaxModShift: Model Privacy via Designed Shifts](https://mlsanity.com/paper/2608.09328/) (2026)

## Abstract (arXiv)

In this paper, shifts are introduced to preserve model privacy against an eavesdropper in federated learning. Model learning is treated as a parameter estimation problem. This perspective allows us to derive the Fisher Information matrix of the model updates from the shifted updates and drive them to singularity, thus posing a hard estimation problem for Eve. The shifts are securely shared with the central server to maintain model accuracy at the server and participating devices. A convergence test is proposed to detect if model updates have been tampered with and we show that our scheme passes this test. Numerical results show that our scheme achieves a higher model shift when compared to a noise injection scheme while requiring a lesser bandwidth secret channel.