privacy-preserving verification

This work addresses a critical limitation in existing certification schemes for encrypted machine learning models, which only verify model behavior on a fixed audit dataset and thus fail to guarantee generalization to new, identically distributed data—rendering them vulnerable to adversarial manipulation. We formally introduce, for the first time, a generalizable security definition tailored to encrypted model certification and expose fundamental assumptions underlying current zero-knowledge proof–based privacy-preserving auditing protocols that do not hold in practical deployments. To bridge this gap, we propose a unified certification framework integrating secure multi-party computation, zero-knowledge proofs, and statistical generalization theory, providing formal guarantees that audit outcomes generalize to real-world data. Empirical evaluation demonstrates that adversaries can achieve over 99% accuracy during audits while degrading true model performance to below 30%; our protocol effectively mitigates such attacks, aligning theoretical assurances with real-world robustness.

Ensuring functional correctness and performance resilience of network protocols under component failures and adversarial attacks remains a significant challenge. Method: This paper proposes a synergistic analysis framework integrating formal verification with attack synthesis. It models protocol behavior using a formal specification language and employs logical predicates, trace analysis, and model checking to achieve closed-loop verification—simultaneously establishing correctness guarantees and automatically generating realistic attack scenarios. Contribution/Results: Diverging from conventional unidirectional verification, our approach innovatively embeds attack-path generation directly into the verification workflow, enabling reproducible and interpretable failure attribution. Experimental evaluation across multiple mainstream network protocols demonstrates substantial improvements in vulnerability detection rates and attack-surface characterization accuracy. The results validate the feasibility and practicality of formal methods for deep, security-critical analysis of complex network protocols.

研究通过调查Tamarin和ProVerif等工具的用户体验，识别了加密协议验证工具在易用性上的障碍，并提出改进设计的具体建议以提高这些工具的可访问性和可用性。

研究提出一种自动密码分析工作流程，通过生成、测试和优化假设来发现密码系统的缺陷。方法包括识别代数映射错误及分布差异，已验证八个已发布构造的失败。

Formal verification of physical-layer security (PLS) protocols under adversarial attacks remains challenging due to the difficulty in rigorously modeling the tight coupling between authentication and confidentiality—capabilities beyond the scope of existing tools like ProVerif. Method: This paper proposes an interactive formal modeling paradigm based on Isabelle/HOL, enabling animated verification and multi-scenario security analysis. Contribution/Results: We uncover the non-intuitive property that authentication can remain intact even when confidentiality is compromised. Comparative validation against ProVerif confirms and strengthens prior confidentiality guarantees. Our framework successfully verifies session-key security of an enhanced PLS-aware Diffie–Hellman protocol under multiple eavesdropping locations and active attacks. The framework integrates watermarking and jamming mechanisms and provides a web interface for practical deployment and analysis.

This study addresses the missing chain of trust in state transition verification for GKR circuit applications by formally verifying sparse Merkle trees and the GKR protocol within Isabelle/HOL, thereby establishing a complete trust chain from the compiler to the proof layer. Methodologically, it achieves end-to-end verification by integrating a Rust executable prover, Creusot/Why3 contracts, and KoalaBear field extension techniques. The core contributions are threefold: first, it unifies compiler correctness and the GKR assembly model within a single proof assistant for the first time, explicitly delineating residual cryptographic obligations; second, it reduces semantic invalidity to explicit event bounds, exposing potential assumption vulnerabilities; and third, it ensures a reliable connection between the implementation code and the formal model.

论文探讨了在证明验证自动化背景下，数学知识确认中人工裁决稀缺的问题，并提出了解决代表忠实性和认知重要性问题的方法。

This work addresses the frequent disconnect between the mathematical certainty of numerical values in cryptographic protocols and their concrete representations, which undermines interoperability and formal verification. Drawing from representation theory, the paper introduces three classes of representations—algorithmically approximable, finitely precisely describable, and canonically normalizable—and proves that no universal computable canonicalizer can transform arbitrary approximate programs into a unique finite encoding. It extends the canonical encoding paradigm of the rational number system Σ_Q to practical cryptographic objects. By integrating computability theory with canonical serialization techniques, the approach is applied to symmetric and asymmetric encryption, hashing, and blockchain integrity protocols. Case studies such as Snaproot demonstrate that canonical representations are essential for achieving precise protocol specifications, ensuring interoperability, and enabling byte-level correctness arguments.

This work addresses the challenge of formally verifying mature, safety-critical industrial C++ codebases by strategically integrating theorem proving (PVS) and model checking (SeaHorn), augmented with large language models to assist in specification construction. The approach is applied to the core order book algorithm of Stellar’s SDEX blockchain module. The verification effort successfully establishes critical correctness properties—including state consistency and unreachability of erroneous states—uncovers discrepancies between documentation and implementation, and produces reusable formal artifacts. These assets enable continuous validation of invariants during future code evolution, thereby enhancing long-term reliability and maintainability of the system.

研究通过引入线性方程基准评估大语言模型在数学验证中的鲁棒性，发现模型对非标准但正确的解题过程敏感，提出监督微调等方法提升鲁棒性。