FuseFSS: New Compiler Improves Speed and Efficiency of Secure LLM Inference
Researchers have developed FuseFSS, a compiler that streamlines secure inference for large language models using function secret sharing, allowing queries without revealing prompts or data. The new approach replaces individual protocols for each operation with a unified compilation pipeline, improving efficiency across nonlinear functions and helper operations. This advancement matters because it makes privacy-preserving AI inference more practical by reducing computational overhead and communication costs while maintaining accuracy.
FuseFSS is a compiler designed to improve the efficiency of two-server secure inference systems that allow clients to query hosted large language models without revealing their prompts or embeddings. Previous GPU-based systems using function secret sharing (FSS) handled linear layers efficiently but struggled with fixed-point nonlinearities and helper operations, each requiring custom protocol implementations. The new compiler unifies this approach by replacing per-operator protocol design with a single compilation pipeline that specifies interval partitions, arithmetic pieces, and predicate bits for each scalar fixed-point operator. Testing on BERT and GPT-style models shows FuseFSS achieves 1.24x to 1.50x end-to-end speedup compared to state-of-the-art FSS-based systems while reducing online communication by 9-16% and lowering preprocessing overhead by 14-23%. The work addresses a key bottleneck in privacy-preserving machine learning by making secure inference more computationally practical.
What's missing
The paper does not discuss potential limitations of the approach, such as scalability to larger models, applicability to other model architectures beyond BERT and GPT-style models, or security assumptions and threat models underlying the two-server inference framework.
What different sources said
- arXiv cs.AICenter
FuseFSS: Efficient Secure LLM Inference with Function Secret Sharing
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.