Preprint · 2025 · Open access

I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference

Gao, Zibo, Hu, Junjie, Guo, Feng, Yixin Zhang, Y. A. Han, Siyuan Liu, Li, Haiyang, Zhiqiang Lv

arXiv (Cornell University) · Published 10 May 2025 · DOI 10.48550/arxiv.2505.06738

Open PDFView at the publisherCite this paper

In brief

“I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference” is a 2025 preprint by Gao, Zibo and 7 co-authors, published in arXiv (Cornell University). OpenAlex counts 13 citations as of 19 November 2026. It is open access: the full text is free to read at arXiv (Cornell University).

Citations
13
Most cited in
2026 (11)
References
41
Authors
8

Abstract

Large Language Models (LLMs) that can be deployed locally have recently gained popularity for privacy-sensitive tasks, with companies such as Meta, Google, and Intel playing significant roles in their development. However, the security of local LLMs through the lens of hardware cache side-channels remains unexplored. In this paper, we unveil novel side-channel vulnerabilities in local LLM inference: token value and token position leakage, which can expose both the victim's input and output text, thereby compromising user privacy. Specifically, we found that adversaries can infer the token values from the cache access patterns of the token embedding operation, and deduce the token positions from the timing of autoregressive decoding phases. To demonstrate the potential of these leaks, we design a novel eavesdropping attack framework targeting both open-source and proprietary LLM inference systems. The attack framework does not directly interact with the victim's LLM and can be executed without privilege. We evaluate the attack on a range of practical local LLM deployments (e.g., Llama, Falcon, and Gemma), and the results show that our attack achieves promising accuracy. The restored output and input text have an average edit distance of 5.2% and 17.3% to the ground truth, respectively. Furthermore, the reconstructed texts achieve average cosine similarity scores of 98.7% (input) and 98.0% (output).

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Zibo, G., Junjie, H., Feng, G., Zhang, Y., Han, Y. A., Liu, S., Haiyang, L., & Lv, Z. (2025). I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2505.06738

MLA 9

Zibo, Gao,, et al. “I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference.” arXiv (Cornell University), 2025. https://doi.org/10.48550/arxiv.2505.06738.

Chicago

Zibo, Gao,, Hu, Junjie, Guo, Feng, Yixin Zhang, Y. A. Han, Siyuan Liu, Li, Haiyang, and Zhiqiang Lv. 2025. “I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference.” arXiv (Cornell University). https://doi.org/10.48550/arxiv.2505.06738.

Harvard

Zibo, G. et al. (2025) ‘I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference’, arXiv (Cornell University). doi:10.48550/arxiv.2505.06738.

IEEE

G. Zibo et al., “I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference,” arXiv (Cornell University), 2025, doi: 10.48550/arxiv.2505.06738.

BibTeX

@misc{zibo2025know,
  title     = {{I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference}},
  author    = {Zibo, Gao, and Junjie, Hu, and Feng, Guo, and Zhang, Yixin and Han, Y. A. and Liu, Siyuan and Haiyang, Li, and Lv, Zhiqiang},
  howpublished = {arXiv (Cornell University)},
  year      = {2025},
  doi       = {10.48550/arxiv.2505.06738},
  url       = {https://doi.org/10.48550/arxiv.2505.06738}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Computer Science › Hardware and Architecture › Security and Verification in Computing

  • local large language models
  • eavesdropping
  • token embeddings
  • autoregressive decoding
  • side-channel vulnerabilities
  • llamas
  • falcons
  • Gemma

Work with this paper

Most-cited papers in Computer Science

All highly cited Computer Science papers

Questions about this paper

How do I cite this paper in APA style?

Zibo, G., Junjie, H., Feng, G., Zhang, Y., Han, Y. A., Liu, S., Haiyang, L., & Lv, Z. (2025). I Know What You Said: Unveiling Hardware Cache Side-Channels in Local Large Language Model Inference. arXiv (Cornell University). https://doi.org/10.48550/arxiv.2505.06738

How many times has this paper been cited?

13 times, according to OpenAlex on 19 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at arXiv (Cornell University): sign in and open the PDF from this page.

Who wrote this paper?

Gao, Zibo, Hu, Junjie, Guo, Feng, Yixin Zhang, Y. A. Han, Siyuan Liu, Li, Haiyang, Zhiqiang Lv.

What is the DOI of this paper?

10.48550/arxiv.2505.06738, permanently at https://doi.org/10.48550/arxiv.2505.06738. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 19 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.