Backend
HE-LRM: Encrypted Deep Learning Recommendation Models using Fully Homomorphic Encryption
arXiv cs.CR0 views
arXiv:2506.18150v5 Announce Type: replace Abstract: FHE inference systems primarily target models with dense inputs, while Deep Learning Recommendation Models (DLRMs) contain sparse categorical features that require lookups into embedding tables. Encrypted lookups are challenging because indexing is not natively supported and must be expressed using SIMD arithmetic. A direct one-hot formulation does not scale. Prior work addresses parts of this problem through embedding compression or compact encrypted queries. However, it does not address complete DLRM inference, leaving open how private lookup interacts with downstream computation and memory constraints. We present HE-LRM, the first end-to-end FHE system for private DLRM inference. We use deterministic compositional embedding compression with client-side digit decomposition, avoiding expensive homomorphic selector construction while consuming only one multiplicative level. We further develop a block-diagonal packing strategy that performs lookups across multiple embedding tables using optimized BSGS linear transformations and produces encrypted embeddings directly compatible with downstream inference. We extend Orion with a high-performance GPU backend and achieve 9 ms for single-inference embedding lookup. We also show that GPU memory capacity limits batching and therefore the latency-bound throughput in such systems. On Criteo, HE-LRM achieves complete encrypted inference with server-side latencies of 154-247 ms on a single GPU. We open-source HE-LRM at https://github.com/baahl-nyu/orion
Source:https://arxiv.org/abs/2506.18150PreviousPropose with the Gradient, Accept with the Loss: One Hedged Attack for Pixel-Sparse RobustnessNextPSM: Prompt Sensitivity Minimization via LLM-Guided Black-Box Optimization
Related posts
Where Is the Cost of Third-Party API Routers in Agentic Software Development?
0arXiv cs.SE
RyBench: Runtime Verification Benchmarking Framework over Publish/Subscribe Networks
0arXiv cs.SE
When the World Is the Test Suite: Agentic Intelligence for the Physical World
0arXiv cs.SE
Direct Numerical Simulation of Thermoacoustically Unstable Flames Via Non-Drifting Acoustic Delay Characteristic Boundary Conditions
0arXiv cs.NA
On the sharp linear convergence rate of the circumcentered--reflection method on subspaces
0arXiv cs.NA
Quantum Gibbs sampling through the detectability lemma
0arXiv cs.NA