mlsanityData to 14 Oct 2026

A Graph-Based Analysis of Comparability, Scale, and Attribution in Private Transformer Inference

arXiv 2610.19607 ↗First posted 14 Oct 2026cs.CR cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Rising privacy concerns around large language models have accelerated research in private Transformer inference (PTI). Synthesizing the empirical claims in this line of research, however, is hindered by multi-dimensional cryptographic progress and inconsistent evaluation contracts. In this work, we systematize the field by constructing an empirical graph of 62 PTI systems, mapping system attributes to node features and evaluation compatibility to edges. A three-tiered analysis of this graph illuminates the state of private inference. Node-level attribution shows that cryptographic progress drives performance gains across most systems, yet always co-occurs with other factors (model fidelity, hardware, cost accounting). Edge-level auditing shows that only 1 of 1891 candidate system pairs is directly comparable under shared assumptions. Finally, graph-level analysis shows that the scaling frontier of private inference has plateaued at 13B parameters and projects the cost of deploying current PTI techniques on production-scale Transformers. Overall, this graph-based framework equips the community to evaluate PTI claims transparently and channel technical speedups into deployment-scale private inference.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 43 of its 87 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

43 of 87 references.

References · 43

  1. 2017Attention Is All You Needcited by 60,361
  2. 2022Efficiently Scaling Transformer Inferencecited by 539
  3. 2024On Protecting the Data Privacy of Large Language Models (LLMs): A Surveycited by 96
  4. 2022MPCFormer: fast, performant and private Transformer inference with MPCcited by 69
  5. 2022THE-X: Privacy-Preserving Transformer Inference with Homomorphic Encryptioncited by 57
  6. 2023PUMA: Secure Inference of LLaMA-7B in Five Minutescited by 48
  7. 2023LLMs Can Understand Encrypted Prompt: Towards Privacy-Computing Friendly Transformerscited by 22
  8. 2022MPCViT: Searching for Accurate and Efficient MPC-Friendly Vision Transformer with Heterogeneous Attentioncited by 20
  9. 2023Primer: Fast Private Transformer Inference on Encrypted Datacited by 20
  10. 2024Encryption-Friendly LLM Architecturecited by 17
  11. 2024Nimbus: Secure and Efficient Two-Party Inference for Transformerscited by 16
  12. 2024PermLLM: Private Inference of Large Language Models within 3 Seconds under WANcited by 14
31 more
  1. 2025CipherPrune: Efficient and Scalable Private Transformer Inferencecited by 10
  2. 2023MERGE: Fast Private Text Generationcited by 9
  3. 2024SecFormer: Fast and Accurate Privacy-Preserving Inference for Transformer Models via SMPCcited by 9
  4. 2025Entropy-Guided Attention for Private LLMscited by 9
  5. 2024Ditto: Quantization-aware Secure Inference of Transformers upon MPCcited by 8
  6. 2024Power-Softmax: Towards Secure LLM Inference over Encrypted Datacited by 7
  7. 2025Breaking the Layer Barrier: Remodeling Private Transformer Inference with Hybrid CKKS and MPCcited by 7
  8. 2023Secure Transformer Inference Protocolcited by 6
  9. 2024MPC-Minimized Secure LLM Inferencecited by 6
  10. 2025Efficient Decoding Methods for Language Models on Encrypted Datacited by 5
  11. 2025CryptoMoE: Privacy-Preserving and Scalable Mixture of Experts Inference via Balanced Expert Routingcited by 5
  12. 2026Cachemir: Fully Homomorphic Encrypted Inference of Generative Large Language Model with KV Cachecited by 5
  13. 2025Comet: Accelerating Private Inference for Large Language Model by Predicting Activation Sparsitycited by 4
  14. 2026What Does the Server See? Understanding Privacy Leakage from Large Language Models in Split Inferencecited by 4
  15. 2021SoK: Privacy-preserving Deep Learning with Homomorphic Encryptioncited by 3
  16. 2024A Survey on Private Transformer Inferencecited by 3
  17. 2025MPCache: MPC-Friendly KV Cache Eviction for Efficient Private LLM Inferencecited by 3
  18. 2026CryptoGen: Secure Transformer Generation with Encrypted KV-Cache Reusecited by 3
  19. 2024TruncFormer: Private LLM Inference Using Only Truncationscited by 2
  20. 2025Private Transformer Inference in MLaaS: A Surveycited by 2
  21. 2025An Efficient Private GPT Never Autoregressively Decodescited by 2
  22. 2025ENSI: Efficient Non-Interactive Secure Inference for Large Language Modelscited by 2
  23. 2026SecMoE: Communication-Efficient Secure MoE Inference via Select-Then-Computecited by 2
  24. 2026On the (In-)Security of the Shuffling Defense in the Transformer Secure Inferencecited by 2
  25. 2026ATLAS: Automated Approximation of Transformers for Efficient Homomorphic Inference in One Hourcited by 2
  26. 2024CipherFormer: Efficient Transformer Private Inference with Low Round Complexitycited by 1
  27. 2025Practical Secure Inference Algorithm for Fine-tuned Large Language Model Based on Fully Homomorphic Encryptioncited by 1
  28. 2025SoK: Can Fully Homomorphic Encryption Support General AI Computation? A Functional and Cost Analysiscited by 1
  29. 2025Cascade: Token-Sharded Private LLM Inferencecited by 1
  30. 2026EncFormer: Secure and Efficient Transformer Inference over Encrypted Datacited by 1
  31. 2026FuseFSS: Efficient Secure LLM Inference with Function Secret Sharingcited by 1

Built on by

Not cited yet.