There is no denying that the use of Information Technology (IT) is undergoing exponential growth in today's world. This digital transformation has also given rise to a multitude of security challenges, notably in the realm of cybercrime. In response to these growing threats, public and private sectors have prioritized the strengthening of IT security measures. In light of the growing security concern, Artificial Intelligence (AI) has gained prominence within the cybersecurity landscape. This paper presents a comprehensive survey of recent advancements in AI-driven threat response systems. To the best of our knowledge, the most recent survey covering the AI reaction domain was conducted in 2017. Since then, considerable literature has been published and therefore it is worth reviewing it. By means of several shared features, each of the studies is compared on a common ground. Through an analysis of the research papers conducted on a standardized basis, this survey aims to unravel the complexities and opportunities of integrating AI into cyber defense. The conclusions drawn from this collective analysis provide a comprehensive snapshot of the evolving landscape at the intersection of AI and cybersecurity. This landscape underscores the growing significance of not only anticipating and detecting threats but also responding to them effectively. Additionally, from these reviews, various research challenges for the future are presented. These challenges serve as a roadmap for researchers and practitioners in the field of AI-integrated reactive strategies.

The rapid development of graph neural networks (GNNs) encourages the rising of link prediction, achieving promising performance with various applications. Unfortunately, through a comprehensive analysis, we surprisingly find that current link predictors with dynamic negative samplers (DNSs) suffer from the migration phenomenon between "easy" and "hard" samples, which goes against the preference of DNS of choosing "hard" negatives, thus severely hindering capability. Towards this end, we propose the MeBNS framework, serving as a general plugin that can potentially improve current negative sampling based link predictors. In particular, we elaborately devise a Meta-learning Supported Teacher-student GNN (MST-GNN) that is not only built upon teacher-student architecture for alleviating the migration between "easy" and "hard" samples but also equipped with a meta learning based sample re-weighting module for helping the student GNN distinguish "hard" samples in a fine-grained manner. To effectively guide the learning of MST-GNN, we prepare a Structure enhanced Training Data Generator (STD-Generator) and an Uncertainty based Meta Data Collector (UMD-Collector) for supporting the teacher and student GNN, respectively. Extensive experiments show that the MeBNS achieves remarkable performance across six link prediction benchmark datasets.

Membership Inference Attacks (MIA) aim to infer whether a target data record has been utilized for model training or not. Prior attempts have quantified the privacy risks of language models (LMs) via MIAs, but there is still no consensus on whether existing MIA algorithms can cause remarkable privacy leakage on practical Large Language Models (LLMs). Existing MIAs designed for LMs can be classified into two categories: reference-free and reference-based attacks. They are both based on the hypothesis that training records consistently strike a higher probability of being sampled. Nevertheless, this hypothesis heavily relies on the overfitting of target models, which will be mitigated by multiple regularization methods and the generalization of LLMs. The reference-based attack seems to achieve promising effectiveness in LLMs, which measures a more reliable membership signal by comparing the probability discrepancy between the target model and the reference model. However, the performance of reference-based attack is highly dependent on a reference dataset that closely resembles the training dataset, which is usually inaccessible in the practical scenario. Overall, existing MIAs are unable to effectively unveil privacy leakage over practical fine-tuned LLMs that are overfitting-free and private. We propose a Membership Inference Attack based on Self-calibrated Probabilistic Variation (SPV-MIA). Specifically, since memorization in LLMs is inevitable during the training process and occurs before overfitting, we introduce a more reliable membership signal, probabilistic variation, which is based on memorization rather than overfitting. Furthermore, we introduce a self-prompt approach, which constructs the dataset to fine-tune the reference model by prompting the target LLM itself. In this manner, the adversary can collect a dataset with a similar distribution from public APIs.

However, current watermark detection algorithms require the secret key used in the watermark generation process, making them susceptible to security breaches and counterfeiting during public detection. To address this limitation, we propose an unforgeable publicly verifiable watermark algorithm that uses two different neural networks for watermark generation and detection, instead of using the same key at both stages. Meanwhile, the token embedding parameters are shared between the generation and detection networks, which makes the detection network achieve a high accuracy very efficiently. Experiments demonstrate that our algorithm attains high detection accuracy and computational efficiency through neural networks with a minimized number of parameters. Subsequent analysis confirms the high complexity involved in forging the watermark from the detection network. Therefore, texts generated by LLMs need to be detected and tagged. At present, some watermarking algorithms for LLM have proved successful in making machinegenerated texts detectable by adding implicit features during the text generation process that are difficult for humans to discover but easily detected by the specially designed method (Christ et al., 2023; Kirchenbauer et al., 2023). The current watermark algorithms for large models utilize a shared key during the generation and detection of watermarks. They work well when the detection access is restricted to the watermark owner only. However, in many situations, when third-party watermark detection is required, the exposure of the shared key would enable others to forge the watermark. Therefore, preventing the watermark forge in the public detection setting, is of great importance. In this work, we propose the first unforgeable publicly verifiable watermarking algorithm for large language models (LLMs).

Split learning enables efficient and privacy-aware training of a deep neural network by splitting a neural network so that the clients (data holders) compute the first layers and only share the intermediate output with the central compute-heavy server. This paradigm introduces a new attack medium in which the server has full control over what the client models learn, which has already been exploited to infer the private data of clients and to implement backdoors in the client models. Although previous work has shown that clients can successfully detect such training-hijacking attacks, the proposed methods rely on heuristics, require tuning of many hyperparameters, and do not fully utilize the clients' capabilities. In this work, we show that given modest assumptions regarding the clients' compute capabilities, an out-of-the-box outlier detection method can be used to detect existing training-hijacking attacks with almost-zero false positive rates. We conclude through experiments on different tasks that the simplicity of our approach we name SplitOut makes it a more viable and reliable alternative compared to the earlier detection methods.

Algorithmic fairness, the research field of making machine learning (ML) algorithms fair, is an established area in ML. As ML technologies expand their application domains, including ones with high societal impact, it becomes essential to take fairness into consideration during the building of ML systems. Yet, despite its wide range of socially sensitive applications, most work treats the issue of algorithmic bias as an intrinsic property of supervised learning, i.e., the class label is given as a precondition. Unlike prior studies in fairness, we propose an individual fairness measure and a corresponding algorithm that deal with the challenges of uncertainty arising from censorship in class labels, while enforcing similar individuals to be treated similarly from a ranking perspective, free of the Lipschitz condition in the conventional individual fairness definition. We argue that this perspective represents a more realistic model of fairness research for real-world application deployment and show how learning with such a relaxed precondition draws new insights that better explains algorithmic fairness. We conducted experiments on four real-world datasets to evaluate our proposed method compared to other fairness models, demonstrating its superiority in minimizing discrimination while maintaining predictive performance with uncertainty present.

The quality of training datasets for deep neural networks is a key factor contributing to the accuracy of resulting models. This effect is amplified in difficult tasks such as object detection. Dealing with errors in datasets is often limited to accepting that some fraction of examples are incorrect, estimating their confidence, and either assigning appropriate weights or ignoring uncertain ones during training. In this work, we propose a different approach. We introduce the Confident Learning for Object Detection (CLOD) algorithm for assessing the quality of each label in object detection datasets, identifying missing, spurious, mislabeled, and mislocated bounding boxes and suggesting corrections. By focusing on finding incorrect examples in the training datasets, we can eliminate them at the root. Suspicious bounding boxes can be reviewed to improve the quality of the dataset, leading to better models without further complicating their already complex architectures. The proposed method is able to point out nearly 80% of artificially disturbed bounding boxes with a false positive rate below 0.1. Cleaning the datasets by applying the most confident automatic suggestions improved mAP scores by 16% to 46%, depending on the dataset, without any modifications to the network architectures. This approach shows promising potential in rectifying state-of-the-art object detection datasets.

Text classification is an important topic in the field of natural language processing. It has been preliminarily applied in information retrieval, digital library, automatic abstracting, text filtering, word semantic discrimination and many other fields. The aim of this research is to use a variety of algorithms to test the ability to identify offensive posts and evaluate their performance against a variety of assessment methods. The motivation for this project is to reduce the harm of these languages to human censors by automating the screening of offending posts. The field is a new one, and despite much interest in the past two years, there has been no focus on the object of the offence. Through the experiment of this project, it should inspire future research on identification methods as well as identification content.

We consider the critical issue of false negatives in Vision-Language Pre-training (VLP), a challenge that arises from the inherent many-to-many correspondence of image-text pairs in large-scale web-crawled datasets. The presence of false negatives can impede achieving optimal performance and even lead to learning failures. To address this challenge, we propose a method called COSMO (COnverting and SMOoothing false negatives) that manages the false negative issues, especially powerful in hard negative sampling. Building upon the recently developed GRouped mIni-baTch sampling (GRIT) strategy, our approach consists of two pivotal components: 1) an efficient connection mining process that identifies and converts false negatives into positives, and 2) label smoothing for the image-text contrastive loss (ITC). Our comprehensive experiments verify the effectiveness of COSMO across multiple downstream tasks, emphasizing the crucial role of addressing false negatives in VLP, potentially even surpassing the importance of addressing false positives. In addition, the compatibility of COSMO with the recent BLIP-family model is also demonstrated.

One of the motivations for explainable AI is to allow humans to make better and more informed decisions regarding the use and deployment of AI models. But careful evaluations are needed to assess whether this expectation has been fulfilled. Current evaluations mainly focus on algorithmic properties of explanations, and those that involve human subjects often employ subjective questions to test human's perception of explanation usefulness, without being grounded in objective metrics and measurements. In this work, we evaluate whether explanations can improve human decision-making in practical scenarios of machine learning model development. We conduct a mixed-methods user study involving image data to evaluate saliency maps generated by SmoothGrad, GradCAM, and an oracle explanation on two tasks: model selection and counterfactual simulation. To our surprise, we did not find evidence of significant improvement on these tasks when users were provided with any of the saliency maps, even the synthetic oracle explanation designed to be simple to understand and highly indicative of the answer. Nonetheless, explanations did help users more accurately describe the models. These findings suggest caution regarding the usefulness and potential for misunderstanding in saliency-based explanations.