mlsanityData to 7 Oct 2026

Communication-Efficient Verifiable Attention for LLM Inference

arXiv 2606.16352 ↗First posted 15 Jun 2026cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Computation integrity of remote large language model (LLM) serving can be questionable. For conventional deep neural networks (DNNs), the existing TEE-shielded DNN partitioning (TSDP) approach uses Trusted Execution Environment (TEE) to compute non-linear components and verify the integrity of linear components offloaded to an untrusted GPU. However, directly applying TSDP to Transformer-based LLMs incurs significant TEE computation and TEE-GPU communication overhead. This paper presents Communication-efficient TEE-GPU Attention (textscVeriAttn) for accelerating verifiable LLM inference. textscVeriAttn offloads both linear and non-linear computations of attention to the GPU, while TEE performs verification. Moreover, for prefill, textscVeriAttn uses a two-level pipeline to overlap data movement, TEE pre-/post-processing, and GPU computation. For decoding, when the key-value cache exceeds available GPU memory, textscVeriAttn partitions attention across TEE and GPU to reduce repeated key-value transfers. Evaluation on an Intel TDX platform shows that textscVeriAttn achieves 2.60-3.38× and 3.86-5.42× acceleration over TSDP for 6k-token prompts and 10k-token outputs during prefill and decoding, respectively.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 32 of its 56 references that are in the map. First posted on arXiv 15 Jun 2026.

Builds on

32 of 56 references.

References · 32

  1. 2017Attention Is All You Needcited by 59,932
  2. 2025Qwen3 Technical Reportcited by 8,195
  3. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  4. 2016Pointer Sentinel Mixture Modelscited by 2,992
  5. 2022GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformerscited by 1,838
  6. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,338
  7. 2024Phi-4 Technical Reportcited by 675
  8. 2021Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LMcited by 621
  9. 2024DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Servingcited by 595
  10. 2023Splitwise: Efficient generative LLM inference using phase splittingcited by 536
  11. 2024Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Servecited by 507
  12. 2024InfiniGen: Efficient Generative Inference of Large Language Models with Dynamic KV Cache Managementcited by 242
20 more
  1. 2023SARATHI: Efficient LLM Inference by Piggybacking Decodes with Chunked Prefillscited by 187
  2. 2025LMCache: An Efficient KV Cache Layer for Enterprise-Scale LLM Inferencecited by 119
  3. 2018Slalom: Fast, Verifiable and Private Execution of Neural Networks in Trusted Hardwarecited by 101
  4. 2024NanoFlow: Towards Optimal Large Language Model Serving Throughputcited by 85
  5. 2024ServerlessLLM: Low-Latency Serverless Inference for Large Language Modelscited by 78
  6. 2024zkLLM: Zero Knowledge Proofs for Large Language Modelscited by 51
  7. 2017SafetyNets: Verifiable Execution of Deep Neural Networks on an Untrusted Cloudcited by 50
  8. 2024FastDecode: High-Throughput GPU-Efficient LLM Serving using Heterogeneous Pipelinescited by 49
  9. 2025KVCache Cache in the Wild: Characterizing and Optimizing KVCache Cache at a Large Cloud Providercited by 41
  10. 2023No Privacy Left Outside: On the (In-)Security of TEE-Shielded DNN Partition for On-Device MLcited by 21
  11. 2025A Survey of Zero-Knowledge Proof Based Verifiable Machine Learningcited by 21
  12. 2020ShadowNet: A Secure and Efficient On-device Model Inference System for Convolutional Neural Networkscited by 20
  13. 2025DeepServe: Serverless Large Language Model Serving at Scalecited by 17
  14. 2025CLONE: Customizing LLMs for Efficient Latency-Aware Inference at the Edgecited by 16
  15. 2024Seq1F1B: Efficient Sequence-Level Pipeline Parallelism for Large Language Model Trainingcited by 11
  16. 2020GOAT: GPU Outsourcing of Deep Learning Training With Asynchronous Probabilistic Integrity Verification Inside Trusted Execution Environmentcited by 6
  17. 2025HGCA: Hybrid GPU-CPU Attention for Long Context LLM Inferencecited by 4
  18. 2024Confidential Computing on Heterogeneous CPU-GPU Systems: Survey and Future Directionscited by 2
  19. 2025Securing Transformer-based AI Execution via Unified TEEs and Crypto-protected Acceleratorscited by 2
  20. 2017Gaussian Variant of Freivalds' Algorithm for Efficient and Reliable Matrix Product Verificationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2023Efficient LLM inference solution on Intel GPU
  2. 2025Efficient LLM Inference with Activation Checkpointing and Hybrid Caching
  3. 2025Make LLM Inference Affordable to Everyone: Augmenting GPU Memory with NDP-DIMM
  4. 2026Characterizing Parallelism Strategies in LLM Inference: Fundamental Compute-Communication Trade-offs
  5. 2026Ragged Paged Attention: A High-Performance and Flexible LLM Inference Kernel for TPU
  6. 2026PackInfer: Compute- and I/O-Efficient Attention for Batched LLM Inference
  7. 2025VecInfer: Efficient LLM Inference with Low-Bit KV Cache via Outlier-Suppressed Vector Quantization
  8. 2025Striking the Right Balance between Compute and Copy: Improving LLM Inferencing Under Speculative Decoding