Pre-trained large deep learning models are now serving as the dominant component for downstream middleware users and have revolutionized the learning paradigm, replacing the traditional approach of training from scratch locally. To reduce development costs, developers often integrate third-party pre-trained deep neural networks (DNNs) into their intelligent software systems. However, utilizing untrusted DNNs presents significant security risks, as these models may contain intentional backdoor defects resulting from the black-box training process. These backdoor defects can be activated by hidden triggers, allowing attackers to maliciously control the model and compromise the overall reliability of the intelligent software. To ensure the safe adoption of DNNs in critical software systems, it is crucial to establish a backdoor defect database for localization studies. This paper addresses this research gap by introducing BDefects4NN, the first backdoor defect database, which provides labeled backdoor-defected DNNs at the neuron granularity and enables controlled localization studies of defect root causes. In BDefects$4 N N$, we define three defect injection rules and employ four representative backdoor attacks across four popular network architectures and three widely adopted datasets, yielding a comprehensive database of$\mathbf{1, 6 5 4}$backdoor-defected DNNs with four defect quantities and varying infected neurons. Based on BDefects4NN, we conduct extensive experiments on evaluating six fault localization criteria and two defect repair techniques, which show limited effectiveness for backdoor defects. Additionally, we investigate backdoor-defected models in practical scenarios, specifically in lane detection for autonomous driving and large language models (LLMs), revealing potential threats and highlighting current limitations in precise defect localization. This paper aims to raise awareness of the threats brought by backdoor defects in our community and inspire future advancements in fault localization methods.

Large Language Models (LLMs) have demonstrated impressive performance across various tasks, yet they remain vulnerable to generating toxic content, necessitating detoxification strategies to ensure safe and responsible deployment. Test-time detoxification methods, which typically introduce static or dynamic interventions into LLM representations, offer a promising solution due to their flexibility and minimal invasiveness. However, current approaches often suffer from imprecise interventions, primarily due to their insufficient exploration of the transition space between toxic and non-toxic outputs. To address this challenge, we propose \textsc{A}utoregressive \textsc{R}eward \textsc{G}uided \textsc{R}epresentation \textsc{E}diting (ARGRE), a novel test-time detoxification framework that explicitly models toxicity transitions within the latent representation space, enabling stable and precise reward-guided editing. ARGRE identifies non-toxic semantic directions and interpolates between toxic and non-toxic representations to reveal fine-grained transition trajectories. These trajectories transform sparse toxicity annotations into dense training signals, enabling the construction of an autoregressive reward model that delivers stable and precise editing guidance. At inference, the reward model guides an adaptive two-step editing process to obtain detoxified representations: it first performs directional steering based on expected reward gaps to shift representations toward non-toxic regions, followed by lightweight gradient-based refinements. Extensive experiments across 8 widely used LLMs show that ARGRE significantly outperforms leading baselines in effectiveness (-62.21\% toxicity) and efficiency (-47.58\% inference time), while preserving the core capabilities of the original model with minimal degradation. Our code is available at the \href{https://anonymous.4open.science/r/ARGRE-6291}{anonymous website}.

Lane detection (LD) is an essential component of autonomous driving systems, providing fundamental functionalities like adaptive cruise control and automated lane centering. Existing LD benchmarks primarily focus on evaluating common cases, neglecting the robustness of LD models against environmental illusions such as shadows and tire marks on the road. This research gap poses significant safety challenges since these illusions exist naturally in real-world traffic situations. For the first time, this paper studies the potential threats caused by these environmental illusions to LD and establishes the first comprehensive benchmark LanEvil for evaluating the robustness of LD against this natural corruption. We systematically design 14 prevalent yet critical types of environmental illusions (e.g., shadow, reflection) that cover a wide spectrum of real-world influencing factors in LD tasks. Based on real-world environments, we create 94 realistic and customizable 3D cases using the widely used CARLA simulator, resulting in a dataset comprising 90,292 sampled images. Through extensive experiments, we benchmark the robustness of popular LD methods using LanEvil, revealing substantial performance degradation (-5.37% Accuracy and -10.70% F1-Score on average), with shadow effects posing the greatest risk (-7.39% Accuracy). Additionally, we assess the performance of commercial auto-driving systems OpenPilot and Apollo through collaborative simulations, demonstrating that proposed environmental illusions can lead to incorrect decisions and potential traffic accidents. To defend against environmental illusions, we propose the Attention Area Mixing (AAM) approach using hard examples, which witness significant robustness improvement (+3.76%) under illumination effects. We hope our paper can contribute to advancing more robust auto-driving systems in the future. Part of our dataset and demos can be found at the https://lanevil.github.io/.

The abuse of deepfake techniques has raised serious concerns about social security and ethical problems, which motivates the development of deepfake detection. However, without fully addressing the domain gap issue, existing deepfake detection methods still show weak generalization ability among datasets belonging to different domains with domain-specific characteristics like identities and generation methods, limiting their practical applications. In this paper, we propose theInvariant Domain-oriented Deepfake Detection method (ID$_{3}$), which improves the generalization of deepfake detection on multiple domains through invariant risk minimization, a novel learning paradigm that addresses the domain gap problem by jointly training a purified invariant predictor and learning an aligned invariant representation. To train a purified invariant predictor, we design theDomain Refinement Data Augmentationstrategy with self-face-swapping and region-erasing approaches, which suppresses domain-specific features and encourages the models to focus on critical domain-invariant characteristics. To learn an aligned invariant representation, we propose theDomain Calibration Batch Normalizationapproach with multiple BN branches, which normalizes input features from different domains into aligned representations during both training and testing. Extensive experiments on multiple datasets demonstrate that our framework can boost the deepfake detection generalization ability and outperform other baselines by large margins. Our codes can be found here

Machine learning (ML) systems have achieved remarkable performance across a wide area of applications. However, they frequently exhibit unfair behaviors in sensitive application domains (e.g., employment and loan), raising severe fairness concerns. To evaluate and test fairness, engineers often generate individual discriminatory instances to expose unfair behaviors before model deployment. However, existing baselines ignore the naturalness of generation and produce instances that deviate from the real data distribution, which may fail to reveal the actual model fairness since these unnatural discriminatory instances are unlikely to appear in practice. To address the problem, this paper proposes a framework named Latent Imitator (LIMI) to generate more natural individual discriminatory instances with the help of a generative adversarial network (GAN), where we imitate the decision boundary of the target model in the semantic latent space of GAN and further samples latent instances on it. Specifically, we first derive a surrogate linear boundary to coarsely approximate the decision boundary of the target model, which reflects the nature of the original data distribution. Subsequently, to obtain more natural instances, we manipulate random latent vectors to the surrogate boundary with a one-step movement, and further conduct vector calculation to probe two potential discriminatory candidates that may be more closely located in the real decision boundary. Extensive experiments on various datasets demonstrate that our LIMI outperforms other baselines largely in effectiveness (×9.42 instances), efficiency (×8.71 speeds), and naturalness (+19.65%) on average. In addition, we empirically demonstrate that retraining on test samples generated by our approach can lead to improvements in both individual fairness (45.67% on IFr and 32.81% on IFo) and group fairness (9.86% on SPD and 28.38% on AOD). Our codes can be found on our website.

Despite the broad application of Machine Learning models as a Service (MLaaS), they are vulnerable to model stealing attacks. These attacks can replicate the model functionality by using the black-box query process without any prior knowledge of the target victim model. Existing stealing defenses add deceptive perturbations to the victim's posterior probabilities to mislead the attackers. However, these defenses are now suffering problems of high inference computational overheads and unfavorable trade-offs between benign accuracy and stealing robustness, which challenges the feasibility of deployed models in practice. To address the problems, this paper proposes Isolation and Induction (InI), a novel and effective training framework for model stealing defenses. Instead of deploying auxiliary defense modules that introduce redundant inference time, InI directly trains a defensive model by isolating the adversary's training gradient from the expected gradient, which can effectively reduce the inference computational cost. In contrast to adding perturbations over model predictions that harm the benign accuracy, we train models to produce uninformative outputs against stealing queries, which can induce the adversary to extract little useful knowledge from victim models with minimal impact on the benign performance. Extensive experiments on several visual classification datasets (e.g., MNIST and CIFAR10) demonstrate the superior robustness (up to 48% reduction on stealing accuracy) and speed (up to 25.4× faster) of our InI over other state-of-the-art methods. Our codes can be found in https://github.com/DIG-Beihang/InI-Model-Stealing-Defense.