Slide decks

One deck per paper in the Transformer + FHE collection, written for a student who knows a little machine learning and no cryptography. Every deck follows the same fifteen-slide shape — the problem in plain words, the one big idea, a worked example with real numbers, the threat model, the results, and what the paper does not solve.

Start here — Why transformers are hard to encrypt

Primer · 16 slides
CKKS, slots, noise, depth, bootstrapping, and the four hard operators. Every other deck assumes this one.

Hands-on — FHE by hand

Interactive · 21 slides
Build a homomorphic cipher, break it, then run a small neural network on encrypted inputs. Every box on the slides is live.

Then: the map

Survey (2024) · SoK (2026)
The survey gives you the taxonomy; the SoK gives you the honest scorecard of what works today.

Reading the colours

the same in every deck
blue = encrypted · grey = plaintext · orange = the expensive step · green = this paper's contribution · red = what leaks
Approach
Status

21 of 62 decks written so far. The rest are listed below and are being added module by module.

Start here

2 decks

Surveys & where the field stands

6 decks

FHE-native, non-interactive inference

12 decks

THE-X — Privacy-Preserving Transformer Inference with HE Chen et al. · 2022 The first attempt to push a whole BERT through homomorphic encryption. FHE-only 18 slides → Converting Transformers to Polynomial Form for Secure Inference Zimerman et al. · 2023 Replace every non-polynomial piece of a transformer, then train the network to like it. FHE-only 20 slides → PowerSoftmax — Towards Secure LLM Inference over Encrypted Data Zimerman et al. · 2024 A softmax replacement designed to be cheap under encryption rather than fitted after the fact. model redesign 20 slides → NEXUS — Secure Transformer Inference Made Non-interactive Zhang et al. · 2024 The landmark non-interactive system — one message each way, client can go offline. FHE-only 21 slides → THOR — Secure Transformer Inference with Homomorphic Encryption Moon et al. · 2024 Careful matrix packing to cut the rotation bill of encrypted attention. FHE-only 20 slides → ELLMo — Packing- and Depth-Aware Encrypted Transformer Inference Guzelhan et al. · 2026 Treats packing layout and depth budget as one joint problem instead of two. FHE-only 19 slides → Euston — Efficient and User-Friendly Secure Transformer Inference Gao et al. · 2026 Non-interactive inference with the client's setup burden deliberately kept small. FHE-only 18 slides → STIP — Non-Interactive Transformer Inference via Compact Packing Wang et al. · 2026 Compact packing that fits more of the model into each ciphertext. FHE-only 19 slides → ATLAS — Automated Approximation of Transformers Xie et al. · 2026 Stops hand-tuning polynomial approximations and searches for them automatically. FHE-only 19 slides → Private LLM Inference with Homomorphic Encryption (tutorial) Lim et al. · 2026 A tutorial walk-through of the whole pipeline — a gentle second stop after the primer. 18 slides → Fully Homomorphic Encryption on Llama 3 Abdennebi et al. · 2026 What happens when you point encrypted inference at a modern open LLM. FHE-only 17 slides → Secure Transformer Inference Protocol Yuan et al. · 2023 A protocol-level view of who holds what — and what you get for a permutation instead of a cipher. 18 slides →

Hybrid HE + MPC

16 decks

Iron — Private Inference on Transformers Hao et al. · 2022 The paper that opened two-party private transformer inference. HE + MPC deck not written yet
BOLT — Privacy-Preserving, Accurate and Efficient Inference Pang et al. · 2023 The hybrid baseline everyone else compares against — and the source of the famous round count. HE + MPC deck not written yet
BumbleBee — Secure Two-party Inference for Large Transformers Lu et al. · 2023 Squeezes the communication cost of the hybrid approach hard. HE + MPC deck not written yet
Nimbus — Secure and Efficient Two-Party Inference for Transformers Li et al. · 2024 Moves work into an offline phase so the online phase is short. HE + MPC deck not written yet
CipherGPT — Secure Two-Party GPT Inference Hou et al. · 2023 The first serious attempt at generative, token-by-token private inference. HE + MPC deck not written yet
CipherFormer — Private Inference with Low Round Complexity Wang & Kuang · 2024 Attacks the round count directly — the thing that makes hybrids painful over a real network. HE + MPC deck not written yet
East — Efficient and Accurate Secure Transformer Framework Ding et al. · 2023 Protocol-level care for the non-linear functions, without giving up accuracy. HE + MPC deck not written yet
Primer — Fast Private Transformer Inference on Encrypted Data Zheng et al. · 2023 Combines several cryptographic tools rather than betting on one. HE + MPC deck not written yet
BLB — Breaking the Layer Barrier Xu et al. · 2025 Stops treating the transformer layer as an atom and re-cuts the HE/MPC boundary inside it. HE + MPC deck not written yet
EncFormer — Secure and Efficient Transformer Inference Zhu et al. · 2026 Chooses where to put each operation by what packing the next one needs. HE + MPC deck not written yet
ENSI — Efficient Non-Interactive Secure Inference for LLMs He et al. · 2025 Pushes a hybrid-style design towards a single round. FHE-only deck not written yet
CipherPrune — Efficient and Scalable Private Transformer Inference Zhang et al. · 2025 Drops tokens that do not matter — the cheapest ciphertext is the one you never compute. HE + MPC deck not written yet
Comet — Private Inference by Predicting Activation Sparsity Yan et al. · 2025 Predicts which activations will be zero and skips them. Fast — and a leak worth studying. HE + MPC deck not written yet
Reducing Key-Switching Overhead in Two-Party Transformer Inference Yang et al. · 2026 A close look at one specific expensive primitive that shows up everywhere. HE + MPC deck not written yet
LLMs Can Understand Encrypted Prompt Liu & Liu · 2023 Reshapes the model to be privacy-computing friendly rather than only the protocol. HE + MPC deck not written yet
Powerformer — High-Accuracy Privacy-Preserving Language Modeling Park et al. · 2025 Chases accuracy rather than speed — what does encryption cost in quality? FHE-only deck not written yet

Pure MPC baselines

5 decks

MPCFormer — Fast, Performant and Private Transformer Inference Li et al. · 2023 Distils the model into an MPC-friendly one. No FHE at all — useful contrast. MPC deck not written yet
PUMA — Secure Inference of LLaMA-7B in Five Minutes Dong et al. · 2023 Three-party MPC at LLaMA scale — the speed FHE is measured against. MPC deck not written yet
SecFormer — Fast and Accurate Private Inference via SMPC Luo et al. · 2024 Keeps accuracy while replacing the awkward non-linear functions. MPC deck not written yet
SHAFT — Secure, Handy, Accurate and Fast Transformer Inference Kei & Chow · 2025 Careful protocol engineering for softmax and GELU under secret sharing. MPC deck not written yet
Sigma — Secure GPT Inference with Function Secret Sharing Gupta et al. · 2023 A different cryptographic tool entirely — function secret sharing, with GPU support. MPC deck not written yet

Training & fine-tuning under encryption

5 decks

HETAL — Privacy-preserving Transfer Learning with HE Lee et al. · 2023 Encrypted fine-tuning that actually converges — the ICML paper and its extended version together. FHE-only deck not written yet
BlindTuner — Encrypted Fine-tuning of Transformers Panzade et al. · 2024 Fine-tunes a vision transformer while the training data stays encrypted. FHE-only deck not written yet
PrivTuner — Homomorphic Encryption with LoRA Li et al. · 2024 Only the small LoRA adapters are trained, so only a little has to be encrypted. FHE-only deck not written yet
Encryption-Friendly LLM Architecture Rho et al. · 2024 Changes the architecture itself so the hard operators are never there to approximate. model redesign deck not written yet
The Inhibitor — ReLU and Addition-Based Attention under TFHE Brännvall & Stoian · 2023 Attention rebuilt out of additions and ReLU, for a scheme that hates multiplication. model redesign deck not written yet

Systems & hardware acceleration

8 decks

EncryptedLLM — LLM Inference via GPU-Accelerated FHE de Castro et al. · 2025 The same algorithms, but on a GPU — where most of the recent speed came from. deck not written yet
Cachemir — FHE Inference of Generative LLMs with KV Cache Yu et al. · 2026 Generation needs a KV cache. Keeping that cache encrypted is its own problem. deck not written yet
CryptoGen — Secure Generation with Encrypted KV-Cache Reuse Zhang et al. · 2026 Reuses encrypted cache entries across tokens instead of recomputing them. deck not written yet
AEGIS — Long-Sequence HE Transformer Inference on Multi-GPU Gong & Ran · 2026 Long inputs break the memory budget; this splits the work across GPUs. deck not written yet
FAME — FPGA Acceleration of Secure Matrix Multiplication Xu et al. · 2025 Custom hardware for the one operation encrypted transformers spend most of their time in. deck not written yet
Chameleon — FHE Scheme Switching Acceleration on GPUs Wang et al. · 2024 Different schemes are good at different things — switching between them costs, and this cuts the cost. deck not written yet
A Scalable Multi-GPU Framework for Encrypted Large-Model Inference Jayashankar et al. · 2025 What it takes to scale encrypted inference past one accelerator. deck not written yet
Dataflow-Oriented Classification of GPU-Accelerated HE Nozaki et al. · 2026 Explains why some GPU HE libraries are fast and others are not. deck not written yet

Medical & federated

7 decks

MedBlindTuner — Fine-tuning on Biomedical Images with FHE Panzade et al. · 2024 Encrypted fine-tuning on medical images — the closest paper to what PREMAL is for. FHE-only deck not written yet
Privacy-Preserving Federated Vision Transformer in Medical AI Amin et al. · 2025 Hospitals train together without sharing patient data, using lightweight HE. deck not written yet
FedShield-LLM — Secure Federated Fine-Tuning of an LLM Mia & Amini · 2025 Federated fine-tuning where the updates themselves stay encrypted. deck not written yet
DictPFL — Private Federated Learning on Encrypted Gradients Xue et al. · 2025 Encrypting every gradient is too slow — this encrypts a compact dictionary instead. deck not written yet
Secure Aggregation using Multiparty Homomorphic Encryption Hosseini et al. · 2025 Nobody holds the whole secret key — the key idea behind federated HE. deck not written yet
Towards Privacy-Preserving Federated Learning using Hybrid HE Costa et al. · 2026 Symmetric encryption on the client, homomorphic work on the server. deck not written yet
Disposable-key-based Image Encryption for Vision Transformer Aso et al. · 2024 Not FHE at all — a cheaper, weaker alternative worth understanding to see what FHE buys. deck not written yet

Beyond transformers

1 deck

No deck matches that filter.