As backdoor attacks become more stealthy and robust, they reveal critical weaknesses in current defense strategies: detection methods often rely on coarse-grained feature statistics, and purification methods typically require full retraining or additional clean models. To address these challenges, we propose DUP (Detection-guided Unlearning for Purification), a unified framework that integrates backdoor detection with unlearning-based purification. The detector captures feature-level anomalies by jointly leveraging class-agnostic distances and inter-layer transitions. These deviations are integrated through a weighted scheme to identify poisoned inputs, enabling more fine-grained analysis. Based on the detection results, we purify the model through a parameter-efficient unlearning mechanism that avoids full retraining and does not require any external clean model. Specifically, we innovatively repurpose knowledge distillation to guide the student model toward increasing its output divergence from the teacher on detected poisoned samples, effectively forcing it to unlearn the backdoor behavior. Extensive experiments across diverse attack methods and language model architectures demonstrate that DUP achieves superior defense performance in detection accuracy and purification efficacy.

Abstract Homomorphic signatures have important potential in cloud computing and data privacy protection, but there are still problems such as low signature efficiency, high overhead, and difficulty in instantiation. To solve these problems, an efficient leveled fully homomorphic signature scheme LFHSS with shortened signature values is constructed. The scheme is based on the GPV framework and the RSIS problem on the NTRU lattice. By using the Fast Fourier Sampling algorithm, it achieves efficient signatures with smaller size. It introduces a homomorphic trapdoor function and designs three basic evaluations: homomorphic addition, multiplication, and scalar multiplication. These operations enable the LFHSS scheme to support homomorphic evaluations of functions consisting of addition, multiplication, and scalar multiplication within a certain circuit depth. Additionally, it is proven to be strongly unforgeable, and the scheme is implemented in software with correctness testing and performance analysis conducted. The experimental results show that the security level of LFHSS is 1.14 times higher than that of BCFL23, and the signature generation speed is 300+ times faster, the homomorphic evaluation speed is 7.8 times faster, and the verification speed is 48K times faster than that of BCFL23. The signature length of LFHSS is only 4.86% of BCFL23. The work in this paper is of great significance to the design and application of homomorphic signatures.

Abstract Matrix operation is one of the most basic and practical operations in statistical analysis and machine learning. The secure matrix operation over homomorphic encryption technology can protect the confidentiality of input data. However, it has not come up with an optimal solution for modern machine learning frameworks, partially due to a lack of efficient matrix computation mechanisms. In this paper, a universal secure matrix multiplication scheme MAT-FHE for any dimension matrix over fully homomorphic encryption technology is designed to realize non-square matrix multiplication, such as $$A_{m\times l}\times B_{l\times n}$$ A m × l × B l × n . The matrix is filled into a hypercube structure and encrypted into a single ciphertext. The number of ciphertext multiplications with the highest computational overhead is reduced through operations such as rotating by rows and columns, ciphertext addition, and multiplication of ciphertext and plaintext. After analysis, it is secure under the CPA model, composable, and supports floating matrix continuous multiplication. The CKKS algorithm of the Microsoft SEAL library is used to implement the matrix multiplication of floating point numbers in any dimension. Shared the computing overhead with SIMD technology and improved the implementation speed. In this paper, the operation time of 16-dimensional matrix multiplication is 4.2253s, which is about 1.5 times faster than the existing best square matrix multiplication scheme. The experimental results show that this method is superior to the existing secure matrix multiplication scheme and can be applied to various secure computing outsourcing and machine learning scenarios.

Abstract The attacker in white-box model has full access to software implementation of a cryptographic algorithm and full control over its execution environment. In order to solve the issues of high storage cost and inadequate security about most current white-box cryptographic schemes, WAS, an improved white-box cryptographic algorithm over AS iteration is proposed. This scheme utilizes the AS iterative structure to construct a lookup table with a five-layer ASASA structure, and the maximum distance separable matrix is used as a linear layer to achieve complete diffusion in a small number of rounds. Attackers can be prevented from recovering the key under black-box model. The length of nonlinear layer S and affine layer A in lookup table is 16 bits, which effectively avoids decomposition attack against the ASASA structure and makes the algorithm possess anti-key extraction security under the white-box model, while WAS possesses weak white-box (32 KB, 112)-space hardness to satisfy anti-code lifting security. WAS has provable security and better storage cost than existing schemes, with the same anti-key extraction security and anti-code lifting security, only 128 KB of memory space is required in WAS, which is only 14% of SPACE-16 algorithm and 33% of Yoroi-16 algorithm.

Blockchain technologies have been applied in many areas, from economics, the internet of things to the industrial internet. In order to solve the issue that the Hyperledger Fabric does not currently support Chinese Commercial Cryptographic (CCC) algorithms, we extended the Blockchain Cryptographic Service Provider (BCCSP) module in the Hyperledger Fabric by upgrading the original BCCSP module to support the CCC algorithms SM2 and SM3. Furthermore, we designed a transaction process by using UBCCSP (Upgraded BCCSP), and a new smart contract also has been presented. After that, an improved consortium blockchain information system based on UBCCSP named UCBIS (Consortium Blockchain Information System based on UBCCSP) is proposed. In the Hyperledger Fabric transaction process, the identity information and transaction data are protected by the SM2 and SM3 algorithms, moreover, SM3 is also used in the construction process of smart contracts. Our smart contracts reduce the total data amount and improve query efficiency. Finally, the information query system based on UBCCSP is implemented. After being tested and analyzed, the average time for every query is only 31.162 ms in the blockchain system, which has better performance and higher query efficiency.