Client-side digit decomposition and block-diagonal packing enable server-side private embedding lookups in FHE, cutting embedding-lookup latency by up to 56x versus CodedHeLUT and enabling end-to-end encrypted DLRM inference in 24-489 seconds on a single CPU.
ngraph-he2: A high- throughput framework for neural network inference on encrypted data
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
HE-LRM: Encrypted Deep Learning Recommendation Models using Fully Homomorphic Encryption
Client-side digit decomposition and block-diagonal packing enable server-side private embedding lookups in FHE, cutting embedding-lookup latency by up to 56x versus CodedHeLUT and enabling end-to-end encrypted DLRM inference in 24-489 seconds on a single CPU.