Alessandro Zirilli, Davide Marincione, Evgenios M. Kornaropoulos +2cs.CR cs.AI
Fully homomorphic encryption (FHE) allows a server to run a language model directly on encrypted user prompts, but current approaches remain prohibitively slow. Ciphertexts natively support only addition, multiplication, and rotation, and multiplications may be composed only to a bounded depth before a costly bootstrapping operation is needed to continue. Every nonlinearity must therefore be approximated by an iterative method, and each iteration uses multiplications. A higher iteration count buys precision but exhausts the available depth faster and triggers more bootstraps, which dominate latency. Existing approaches fix the iteration counts uniformly across the model rather than tailoring them to each site's error tolerance. We introduce Homomorphic Encryption-Aware Training (HEAT), a fine-tuning method that makes the per-nonlinearity iteration counts learnable, enabling them and the model weights to co-adapt during training. HEAT optimizes iterations with respect to the task objective, allowing the model to adapt to approximation errors encountered during inference without architectural changes or retraining from scratch. On encrypted GPT-2 decoding, HEAT reduces iterations by $3.1\times$, bootstraps by $1.6\times$, and end-to-end latency by $1.4\times$, while improving decode agreement over the calibrated baseline.
Fully homomorphic encryption (FHE) lets a server run inference on encrypted data with strong privacy guarantees, but running a Transformer under FHE is expensive. Its non-linear operations, such as softmax, normalization, and activation, must be replaced with polynomial approximations that the CKKS scheme supports, and the depth of these approximations dominates inference cost. Existing FHE Transformers use hand-tuned approximation settings, such as iteration count and polynomial degree, applied uniformly across layers, models, and tasks. Hand-tuning is slow and error-prone. Even a single uniform setting has about $10^7$ choices, and manual search cannot exploit layer-wise variation. AutoFHE, the only automated method with multi-objective search, targets ReLU-only CNNs and needs full fine-tuning per candidate, which is too costly for Transformers. Per-layer settings also push the search space to about $10^{85}$ for BERT and ViT and $10^{228}$ for LLaMA3, beyond both manual and fine-tuning-based search. We present ATLAS, a training-free framework that automates this search by treating each layer's approximation setting as a multi-objective optimization over latency and accuracy. The problem is hard: the decision space is large (96 or 256 variables), each configuration takes 70 to 1,000 seconds to evaluate even in cleartext, and 85 to 90 percent of configurations are invalid. ATLAS handles this with a two-stage optimization strategy and a surrogate model, completing the search in about one hour. Compared to an iterative softmax baseline, ATLAS cuts multiplicative depth and end-to-end latency by about 35 percent with little accuracy loss, and works across encoder-only, decoder-only, and vision Transformers, complementing parallel work on packing and matrix multiplication.
Structured pruning is essential for making neural network inference feasible under homomorphic encryption (HE), yet its impact on model reliability has remained unexplored. This paper presents a systematic reliability characterization of pruned CKKS-encrypted neural networks and introduces Polynomial-Sensitivity-Aware Pruning (PSAP), a structured pruning method that is inherently reliability-aware. PSAP scores filters jointly by weight magnitude, polynomial activation sensitivity, and rotation cost, which concentrates pruning in fault-tolerant regions. Across two architectures, two datasets, two numerical representations, and five bit-error rates (40 full-model and 108 per-layer experiments), PSAP-pruned models limit catastrophic (>10 pp accuracy drop) layers to at most two versus 5--14 for magnitude-pruned baselines, reducing worst-case vulnerability by up to 29 times under int32 bit-flip injection. Direct CKKS encrypted fault injection indicates a safe operating boundary near BER~ 10^{-5}, supporting int32 injection as a conservative reliability proxy. The fault-critical structural layers account for only 1.1% of parameters, enabling selective hardening at minimal overhead. These reliability gains are obtained alongside competitive efficiency: PSAP reduces Halevi--Shoup rotations by up to 45.2\% on ResNet-32, and an adaptive mixed-degree allocation scheme lowers multiplicative depth from 66 to 56 levels, enabling leveled inference without bootstrapping.
Fully homomorphic encryption (FHE) enables computation on encrypted data, but practical encrypted Transformer inference is bottlenecked by the sequential composition of many nonlinear blocks. We study whether Structured Newton Layer Parallelism (SNLP) can make this inter-layer composition more FHE-friendly: each Transformer block still requires polynomial approximations for operations such as softmax and RMSNorm, but SNLP reduces the layerwise sequential nonlinear depth from L stages to a small number of solver iterations plus linear structured corrections. Using a simulation framework based on Chebyshev polynomial approximations, we measure error accumulation under sequential versus SNLP inference across 8 models and 4 architecture families. On a 0.5B IDN-trained model, SNLP reduces symbolic bootstraps from 53 to 20 (2.65x) with only +1.2% perplexity degradation, while lowering error amplification (1.36x vs. 1.42x). Across all tested models, SNLP has lower amplification than sequential inference. Ablations show that softmax approximation dominates the error budget and CKKS arithmetic noise is negligible in our setting, suggesting that SNLP is complementary to block-level FHE-friendly operator design rather than a replacement for it.
Fully Homomorphic Encryption (FHE) enables privacy-preserving machine learning but incurs extreme computational and memory overhead. These costs come not only from expensive low-level primitives, including Number Theoretic Transform (NTT), rotation, and key-switching, but also from inefficient ciphertext packing at the application level. Existing packing strategies typically preserve either neighboring data elements or feature grouping, but not both, leading to wasted ciphertext slots, excessive rotations, and inflated ciphertext counts. We propose FEnc2, a unified and principled fragment-based encoding framework for CKKS-based private convolutional neural network inference. FEnc2 optimizes slot utilization, rotation complexity, and ciphertext density through two components: 1)Conv-aware Encoding, which analytically selects an optimal fragment size to decouple spatial dependencies and jointly minimize inner-outer rotations across layers, and 2)Arch-aware Ct Compression, which restores ciphertext density after feature- or channel-reduction layers. Together, these transformations reshape encrypted workload structure and reduce homomorphic operations by one to two orders of magnitude. With full memory capacity utilized, i.e., at maximum batch size, FEnc2 achieves end-to-end latency speedups over the state-of-the-art Orion of up to 228.83x on GPU and 226.06x on CPU for LeNet on MNIST, and up to 4.55x on GPU and 9.43x on CPU for MobileNet on ImageNet. FEnc2 is hardware-agnostic yet architecturally transformative: by optimizing encrypted tensor layout before execution, it reduces ciphertext count and workload pressure on hardware, complementing primitive-level optimizations such as NTT and keyswitch accelerators. These results show that application-level data layout is a first-order architectural design dimension for encrypted inference and an important enabler for next-generation FHE systems.
Fully homomorphic encryption (FHE) supports only additions and multiplications, so FHE-only neural-network inference typically replaces ReLU with polynomials fitted over empirical activation intervals. Such interval fitting often requires higher-degree polynomials to control activation error, incurring homomorphic evaluation costs, while classification is determined by the final logit decision. We revisit ReLU replacement from a decision-aware perspective: given a trained single-hidden-layer ReLU MLP and a specified calibration set, can an HE-friendly low-degree polynomial replace ReLU without retraining while preserving calibration-set decisions? We focus on quadratic replacement, the lowest-degree choice that retains a genuine per-unit nonlinearity. For calibration sets positive-margin separable in the lifted space, we formulate quadratic replacement as a linear separation problem, yielding necessary and sufficient conditions for calibration-lossless replacement and a constructive algorithm for the coefficients. When the positive-margin condition fails -- typically due to a few misclassified calibration samples -- we extend the same geometric framework via reduced convex hulls and Lagrangian-dual soft-margin relaxations, which bound the influence of any single sample, converting the problem into smaller convex quadratic programs that yield approximately feasible coefficients with high empirical agreement on calibration-set decisions. In particular, at the maximal weight cap $μ=1$, the reduced-convex-hull relaxation reduces to the convex-hull separation of the strictly separable case; the relaxation thus continuously extends the exact theory. Under CKKS, the quadratic replacement matches plaintext top-1 accuracy on multiple benchmarks, running 3.7--4.1$\times$ faster than Remez-7 in the activation module and 1.18--1.68$\times$ faster end-to-end.