TIGER delivers the first GPU-accelerated high-precision TFHE implementations for LLM nonlinear layers, with measured speedups of 7.17x for GELU, 16.68x for Softmax, and 17.05x for LayerNorm over CPU baselines.
TFHE-rs: A Pure Rust Implementation of the TFHE Scheme for Boolean and Integer Arithmetics Over Encrypted Data,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
GPU Acceleration of TFHE-Based High-Precision Nonlinear Layers for Encrypted LLM Inference
TIGER delivers the first GPU-accelerated high-precision TFHE implementations for LLM nonlinear layers, with measured speedups of 7.17x for GELU, 16.68x for Softmax, and 17.05x for LayerNorm over CPU baselines.