Arxiv 2509 25072 Optimizing Privacy Preserving Primitives To Support Llm Scal
ASecurity
We present an overview of efforts to bridge the gap between overhead and practicality for privacy-preserving learning systems using MPC, ZKPs, and FHE. Through hardware/software/algorithm co-design, we show progress towards enabling LLM-scale applications in privacy-preserving settings including DNN IP ownership, ethical LLM usage enforcement, and transformer inference.
Installs into .claude/skills of the current project.
Are you the author of Arxiv 2509 25072 Optimizing Privacy Preserving Primitives To Support Llm Scal?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-arxiv-2509-25072-optimizing-privacy-preserving-pri)
---
title: "Optimizing Privacy-Preserving Primitives to Support LLM-Scale Applications"
authors: "Yaman Jandali, Ruisi Zhang, Nojan Sheybani, Farinaz Koushanfar"
arxiv_id: "2509.25072"
categories: "cs.CR; cs.AI; cs.LG"
utility: 0.88
date_added: "2026-09-29"
category: "security-privacy"
---
# Optimizing Privacy-Preserving Primitives to Support LLM-Scale Applications
## Abstract
We present an overview of efforts to bridge the gap between overhead and practicality for privacy-preserving learning systems using MPC, ZKPs, and FHE. Through hardware/software/algorithm co-design, we show progress towards enabling LLM-scale applications in privacy-preserving settings including DNN IP ownership, ethical LLM usage enforcement, and transformer inference.
## Key Contributions
- Novel approach in security privacy domain
- Utility score: 0.88
- Published on arXiv: 2509.25072
## Potential Applications
- Research reference for security privacy
- Building block for related systems
