arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2511.19902cs.CRcs.AI
收藏

基于零知识证明的可验证模型推断

Zero-Knowledge Proof Based Verifiable Inference of Models

Zhejiang University(浙江大学)

Yunxiao Wang

AI总结:

本文提出一种基于零知识证明的框架,实现无需暴露模型参数的深度学习推断验证,通过递归组合零知识证明和Fiat-Shamir启发式方法,生成固定大小的zkSNARK证明,验证DeepSeek模型的可验证性。

摘要:

近年来,人工智能(AI),特别是深度学习,在各种应用中得到了广泛应用。然而,仍然存在一个根本性挑战:当模型所有者无法(或不愿)披露其参数时,如何验证AI模型推断的正确性?这些参数代表着巨大的训练成本和宝贵的知识产权,使透明验证变得困难。在本文中,我们介绍了一个零知识框架,能够在不暴露模型内部参数的情况下验证深度学习推断。该框架基于递归组合的零知识证明,并且不需要可信设置,支持线性和非线性神经网络层,包括矩阵乘法、归一化、softmax和SiLU。利用Fiat-Shamir启发式方法,我们获得了一个简洁的非交互知识论证(zkSNARK),具有固定大小的证明。为了证明我们方法的实用性,我们将DeepSeek模型转换为完全SNARK可验证的版本,称为ZK-DeepSeek,并通过实验表明,我们的框架在现实世界的AI验证工作负载中实现了效率和灵活性。

英文摘要:

Recent advances in artificial intelligence (AI), particularly deep learning, have led to widespread adoption across various applications. Yet, a fundamental challenge persists: how can we verify the correctness of AI model inference when model owners cannot (or will not) reveal their parameters? These parameters represent enormous training costs and valuable intellectual property, making transparent verification difficult. In this paper, we introduce a zero-knowledge framework capable of verifying deep learning inference without exposing model internal parameters. Built on recursively composed zero-knowledge proofs and requiring no trusted setup, our framework supports both linear and nonlinear neural network layers, including matrix multiplication, normalization, softmax, and SiLU. Leveraging the Fiat-Shamir heuristic, we obtain a succinct non-interactive argument of knowledge (zkSNARK) with constant-size proofs. To demonstrate the practicality of our approach, we translate the DeepSeek model into a fully SNARK-verifiable version named ZK-DeepSeek and show experimentally that our framework delivers both efficiency and flexibility in real-world AI verification workloads.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑