22 December 2022 in IEEE Transactions on Evolutionary Computation
Due to the inherent vulnerability of Deep Neural Networks (DNNs), Adversarial Example (AE) attack has become a serious threat to intelligent systems, e.g., the failure cause of an image classification system. Different to existing works, in this paper we are interested in the generation of AEs for DNNs with defensive mechanisms. To make the attack more practical, we exploit a query-based method to generate image AEs in a black-box attack setting. Considering that the generation of AEs is inherently a constrained optimization problem, this paper first formulates three objectives regarding to defensive DNNs, i.e, attack effectiveness, attack evasiveness and attack coverage. Then, this paper proposes a query-efficient AE attack based on Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to address the perturbation optimization problem. To improve the efficiency of search and query, AE-specific operators including block-level and pixel-level crossovers, discrete perturbation mutation and direction-driven reproduction are designed within the GA-based search framework. In addition, predication-based adaptation of reproduction-related parameters is implemented to speed up the search convergence. PSO-based jumping process is further devised to avoid stuck in local optimum. Benchmark-based experiments evaluated the efficiency of our method, which can achieve an attack success rate of 100% with averagely 52.95% reduced queries in contrast to existing black-box attacks on non-defensive models. For defensive DNN models, our method can obtain top attack performance with the query reduction up to 70.92% comparing with the candidates.
1 June 2026 in IEEE Transactions on Evolutionary Computation
Despite the popularity of graph neural networks, perturbed graph data is still a serious threat towards its inherent vulnerabilities. Adversarial examples can still easily manipulate the output of graph neural networks across various attack scenarios. Meanwhile, attacks on graph networks also appear to be crucial, as it can help model designers enhance the robustness of their models. In this study, we propose a sensitivity-oriented poisoning attack for self-supervised graph embedding models through bilevel optimization, which employs different optimization methods at each level. In addition, in order to improve attack effectiveness, we analyze graph structure to identify sensitive nodes and edges that guide attack directions, combining gradient-based and query-based methods to target both edge connections and node attributes. Besides, according to the defects of existing graph masked auto-encoders models, we design the feature sensitivity and feature variance to reduce the feature differentiability, which impairs the performance of the downstream model. Ablation studies validate our operator is effective on three citation datasets. And benchmark-based experiments support the effectiveness of our method on three different graph tasks. Specifically, our approach can achieve an average reduction of 3% in the accuracy of node classification compared to existing methods for attacking neural structures alone. For attacking both graph structures and attributes, our model has even achieved an average reduction of 4.5% for the node classification task, outperforming the existing methods.
26 January 2026 in IEEE Transactions on Evolutionary Computation
Large Language Models (LLMs) demonstrate promising capabilities in natural language understanding and reasoning with enormous parameter spaces and vast amounts of training data. These attributes have facilitated their deployment into diverse application domains. However, the underlying parameters implicitly assume decision-making boundaries, resulting in a significant number of decision spaces not covered by training data. This makes them susceptible to adversarial manipulations through carefully crafted inputs. To illuminate the vulnerabilities of LLMs, we propose a unified token-level jailbreaking attack that makes victim models generate responses for potentially harmful queries. Specifically, we propose an evolutionary algorithm to evolve perturbation sets, utilizing gradient-based and crossover-based operators to enhance performance under multiple scenarios. Furthermore, we develop a repository for reusing past perturbations and conduct an analysis of token sensitivity within LLMs, facilitating zero-shot attacks with convergence. Extensive benchmark experiments validate the effectiveness of our method on three different models, achieving increases of 62.36%, 57.89%, and 64.81% in attack success rates compared to baseline methods under white-box scenarios. In addition, the evaluation experiments demonstrate our method is effective for multiple scenarios and different size of models. This research reveals vulnerabilities of LLMs, provides theoretical foundations for developing more robust defense strategies, and contributes to building more reliable AI systems.
1 January 2026 in IEEE Access
7 April 2025 in Knowledge-Based Systems
7 November 2026 in IEEE Transactions on Mobile Computing