cs.CRJun 8, 2026

FuseFSS: Efficient Secure LLM Inference with Function Secret Sharing

Authors: Yuhan MaYong LiStefan Schmid

Abstract

Two-server secure inference allows a client to query a hosted large language model (LLM) without revealing prompts or embeddings. Recent GPU systems based on function secret sharing (FSS) make linear layers efficient, but fixed-point nonlinearities and helper operations remain a bottleneck because each operator is typically implemented as a bespoke protocol with its own comparisons, wrap-around corrections, and preprocessing material. We present FuseFSS, a compiler that replaces per-operator protocol design with a single compilation pipeline. For each scalar fixed-point operator, a compact specification lists its interval partition, low-degree arithmetic pieces, and required predicate bits. The compiler emits two batched FSS evaluations on the public masked value: one packed comparison that returns all predicate bits, and one vector interval lookup that returns the active coefficients and constants. Compared to the current state-of-the-art FSS-based GPU secure inference, FuseFSS preserves accuracy while achieving a 1.24×1.24\times--1.50×1.50\times end-to-end speedup and reducing online communication by 9%9\%--16%16\% on BERT and GPT-style models; preprocessing is also lighter, with 14%14\%--23%23\% lower key-generation time and 20%20\%--24%24\% smaller keys.

Explore similar work

CardsList