CLIBE: Detecting Dynamic Backdoors in Transformer-based NLP Models

Backdoors can be injected into NLP models to induce misbehavior when the input text contains a specific feature, known as a trigger, which the attacker secretly selects. Unlike fixed words, phrases, or sentences used in the static text trigger, NLP dynamic backdoor attacks design triggers associated with abstract and latent text features, making them considerably stealthier than traditional static backdoor attacks. However, existing research on NLP backdoor detection primarily focuses on defending against static backdoor attacks, while detecting dynamic backdoors in NLP models remains largely unexplored. This paper presents CLIBE, the first framework to detect dynamic backdoors in Transformer-based NLP models. CLIBE injects a "few-shot perturbation" into the suspect Transformer model by crafting optimized weight perturbation in the attention layers to make the perturbed model classify a limited number of reference samples as a target label. Subsequently, CLIBE leverages the generalization ability of this few-shot perturbation to determine whether the original model contains a dynamic backdoor. Extensive evaluation on three advanced NLP dynamic backdoor attacks, two widely-used Transformer frameworks, and four real-world classification tasks strongly validates the effectiveness of CLIBE. We also demonstrate the robustness of CLIBE against various adaptive attacks. Furthermore, we employ CLIBE to scrutinize 49 popular Transformer models on Hugging Face and discover one exhibiting a high probability of containing a dynamic backdoor. We have contacted Hugging Face and provided detailed evidence of this model's backdoor behavior. Moreover, we extend CLIBE to detect backdoor text generation models modified to exhibit toxic behavior. To the best of our knowledge, CLIBE is the first framework capable of detecting backdoors in text generation models without access to trigger input test samples.

SUB-PLAY: Adversarial Policies against Partially Observed Multi-Agent Reinforcement Learning Systems

Recent advances in multi-agent reinforcement learning (MARL) have opened up vast application prospects, including swarm control of drones, collaborative manipulation by robotic arms, and multi-target encirclement. However, potential security threats during the MARL deployment need more attention and thorough investigation. Recent researches reveal that an attacker can rapidly exploit the victim's vulnerabilities and generate adversarial policies, leading to the victim's failure in specific tasks. For example, reducing the winning rate of a superhuman-level Go AI to around 20%. They predominantly focus on two-player competitive environments, assuming attackers possess complete global state observation. In this study, we unveil, for the first time, the capability of attackers to generate adversarial policies even when restricted to partial observations of the victims in multi-agent competitive environments. Specifically, we propose a novel black-box attack (SUB-PLAY), which incorporates the concept of constructing multiple subgames to mitigate the impact of partial observability and suggests the sharing of transitions among subpolicies to improve the exploitative ability of attackers. Extensive evaluations demonstrate the effectiveness of SUB-PLAY under three typical partial observability limitations. Visualization results indicate that adversarial policies induce significantly different activations of the victims' policy networks. Furthermore, we evaluate three potential defenses aimed at exploring ways to mitigate security threats posed by adversarial policies, providing constructive recommendations for deploying MARL in competitive environments.

The Risk of Federated Learning to Skew Fine-Tuning Features and Underperform Out-of-Distribution Robustness

To tackle the scarcity and privacy issues associated with domain-specific datasets, the integration of federated learning in conjunction with fine-tuning has emerged as a practical solution. However, our findings reveal that federated learning has the risk of skewing fine-tuning features and compromising the out-of-distribution robustness of the model. By introducing three robustness indicators and conducting experiments across diverse robust datasets, we elucidate these phenomena by scrutinizing the diversity, transferability, and deviation within the model feature space. To mitigate the negative impact of federated learning on model robustness, we introduce GNP, a \underline{G}eneral \underline{N}oisy \underline{P}rojection-based robust algorithm, ensuring no deterioration of accuracy on the target distribution. Specifically, the key strategy for enhancing model robustness entails the transfer of robustness from the pre-trained model to the fine-tuned model, coupled with adding a small amount of Gaussian noise to augment the representative capacity of the model. Comprehensive experimental results demonstrate that our approach markedly enhances the robustness across diverse scenarios, encompassing various parameter-efficient fine-tuning methods and confronting different levels of data heterogeneity.

MEAOD: Model Extraction Attack against Object Detectors

The widespread use of deep learning technology across various industries has made deep neural network models highly valuable and, as a result, attractive targets for potential attackers. Model extraction attacks, particularly query-based model extraction attacks, allow attackers to replicate a substitute model with comparable functionality to the victim model and present a significant threat to the confidentiality and security of MLaaS platforms. While many studies have explored threats of model extraction attacks against classification models in recent years, object detection models, which are more frequently used in real-world scenarios, have received less attention. In this paper, we investigate the challenges and feasibility of query-based model extraction attacks against object detection models and propose an effective attack method called MEAOD. It selects samples from the attacker-possessed dataset to construct an efficient query dataset using active learning and enhances the categories with insufficient objects. We additionally improve the extraction effectiveness by updating the annotations of the query dataset. According to our gray-box and black-box scenarios experiments, we achieve an extraction performance of over 70% under the given condition of a 10k query budget.

Improving the Robustness of Transformer-based Large Language Models with Dynamic Attention

Transformer-based models, such as BERT and GPT, have been widely adopted in natural language processing (NLP) due to their exceptional performance. However, recent studies show their vulnerability to textual adversarial attacks where the model's output can be misled by intentionally manipulating the text inputs. Despite various methods that have been proposed to enhance the model's robustness and mitigate this vulnerability, many require heavy consumption resources (e.g., adversarial training) or only provide limited protection (e.g., defensive dropout). In this paper, we propose a novel method called dynamic attention, tailored for the transformer architecture, to enhance the inherent robustness of the model itself against various adversarial attacks. Our method requires no downstream task knowledge and does not incur additional costs. The proposed dynamic attention consists of two modules: (I) attention rectification, which masks or weakens the attention value of the chosen tokens, and (ii) dynamic modeling, which dynamically builds the set of candidate tokens. Extensive experiments demonstrate that dynamic attention significantly mitigates the impact of adversarial attacks, improving up to 33\% better performance than previous methods against widely-used adversarial attacks. The model-level design of dynamic attention enables it to be easily combined with other defense methods (e.g., adversarial training) to further enhance the model's robustness. Furthermore, we demonstrate that dynamic attention preserves the state-of-the-art robustness space of the original model compared to other dynamic modeling methods.

Facial Data Minimization: Shallow Model as Your Privacy Filter

Face recognition service has been used in many fields and brings much convenience to people. However, once the user's facial data is transmitted to a service provider, the user will lose control of his/her private data. In recent years, there exist various security and privacy issues due to the leakage of facial data. Although many privacy-preserving methods have been proposed, they usually fail when they are not accessible to adversaries' strategies or auxiliary data. Hence, in this paper, by fully considering two cases of uploading facial images and facial features, which are very typical in face recognition service systems, we proposed a data privacy minimization transformation (PMT) method. This method can process the original facial data based on the shallow model of authorized services to obtain the obfuscated data. The obfuscated data can not only maintain satisfactory performance on authorized models and restrict the performance on other unauthorized models but also prevent original privacy data from leaking by AI methods and human visual theft. Additionally, since a service provider may execute preprocessing operations on the received data, we also propose an enhanced perturbation method to improve the robustness of PMT. Besides, to authorize one facial image to multiple service models simultaneously, a multiple restriction mechanism is proposed to improve the scalability of PMT. Finally, we conduct extensive experiments and evaluate the effectiveness of the proposed PMT in defending against face reconstruction, data abuse, and face attribute estimation attacks. These experimental results demonstrate that PMT performs well in preventing facial data abuse and privacy leakage while maintaining face recognition accuracy.

TextDefense: Adversarial Text Detection based on Word Importance Entropy

Currently, natural language processing (NLP) models are wildly used in various scenarios. However, NLP models, like all deep models, are vulnerable to adversarially generated text. Numerous works have been working on mitigating the vulnerability from adversarial attacks. Nevertheless, there is no comprehensive defense in existing works where each work targets a specific attack category or suffers from the limitation of computation overhead, irresistible to adaptive attack, etc. In this paper, we exhaustively investigate the adversarial attack algorithms in NLP, and our empirical studies have discovered that the attack algorithms mainly disrupt the importance distribution of words in a text. A well-trained model can distinguish subtle importance distribution differences between clean and adversarial texts. Based on this intuition, we propose TextDefense, a new adversarial example detection framework that utilizes the target model's capability to defend against adversarial attacks while requiring no prior knowledge. TextDefense differs from previous approaches, where it utilizes the target model for detection and thus is attack type agnostic. Our extensive experiments show that TextDefense can be applied to different architectures, datasets, and attack methods and outperforms existing methods. We also discover that the leading factor influencing the performance of TextDefense is the target model's generalizability. By analyzing the property of the target model and the property of the adversarial example, we provide our insights into the adversarial attacks in NLP and the principles of our defense method.

All You Need Is Hashing: Defending Against Data Reconstruction Attack in Vertical Federated Learning

Vertical federated learning is a trending solution for multi-party collaboration in training machine learning models. Industrial frameworks adopt secure multi-party computation methods such as homomorphic encryption to guarantee data security and privacy. However, a line of work has revealed that there are still leakage risks in VFL. The leakage is caused by the correlation between the intermediate representations and the raw data. Due to the powerful approximation ability of deep neural networks, an adversary can capture the correlation precisely and reconstruct the data. To deal with the threat of the data reconstruction attack, we propose a hashing-based VFL framework, called \textit{HashVFL}, to cut off the reversibility directly. The one-way nature of hashing allows our framework to block all attempts to recover data from hash codes. However, integrating hashing also brings some challenges, e.g., the loss of information. This paper proposes and addresses three challenges to integrating hashing: learnability, bit balance, and consistency. Experimental results demonstrate \textit{HashVFL}'s efficiency in keeping the main task's performance and defending against data reconstruction attacks. Furthermore, we also analyze its potential value in detecting abnormal inputs. In addition, we conduct extensive experiments to prove \textit{HashVFL}'s generalization in various settings. In summary, \textit{HashVFL} provides a new perspective on protecting multi-party's data security and privacy in VFL. We hope our study can attract more researchers to expand the application domains of \textit{HashVFL}.

Hijack Vertical Federated Learning Models with Adversarial Embedding

Vertical federated learning (VFL) is an emerging paradigm that enables collaborators to build machine learning models together in a distributed fashion. In general, these parties have a group of users in common but own different features. Existing VFL frameworks use cryptographic techniques to provide data privacy and security guarantees, leading to a line of works studying computing efficiency and fast implementation. However, the security of VFL's model remains underexplored.

"Is your explanation stable?": A Robustness Evaluation Framework for Feature Attribution

Understanding the decision process of neural networks is hard. One vital method for explanation is to attribute its decision to pivotal features. Although many algorithms are proposed, most of them solely improve the faithfulness to the model. However, the real environment contains many random noises, which may leads to great fluctuations in the explanations. More seriously, recent works show that explanation algorithms are vulnerable to adversarial attacks. All of these make the explanation hard to trust in real scenarios. To bridge this gap, we propose a model-agnostic method \emph{Median Test for Feature Attribution} (MeTFA) to quantify the uncertainty and increase the stability of explanation algorithms with theoretical guarantees. MeTFA has the following two functions: (1) examine whether one feature is significantly important or unimportant and generate a MeTFA-significant map to visualize the results; (2) compute the confidence interval of a feature attribution score and generate a MeTFA-smoothed map to increase the stability of the explanation. Experiments show that MeTFA improves the visual quality of explanations and significantly reduces the instability while maintaining the faithfulness. To quantitatively evaluate the faithfulness of an explanation under different noise settings, we further propose several robust faithfulness metrics. Experiment results show that the MeTFA-smoothed explanation can significantly increase the robust faithfulness. In addition, we use two scenarios to show MeTFA's potential in the applications. First, when applied to the SOTA explanation method to locate context bias for semantic segmentation models, MeTFA-significant explanations use far smaller regions to maintain 99\%+ faithfulness. Second, when tested with different explanation-oriented attacks, MeTFA can help defend vanilla, as well as adaptive, adversarial attacks against explanations.