Prompting the Unseen: Detecting Hidden Backdoors in Black-Box Models

Visual prompting (VP) is a new technique that adapts well-trained frozen models for source domain tasks to target domain tasks. This study examines VP's benefits for black-box model-level backdoor detection. The visual prompt in VP maps class subspaces between source and target domains. We identify a misalignment, termed class subspace inconsistency, between clean and poisoned datasets. Based on this, we introduce \textsc{BProm}, a black-box model-level detection method to identify backdoors in suspicious models, if any. \textsc{BProm} leverages the low classification accuracy of prompted models when backdoors are present. Extensive experiments confirm \textsc{BProm}'s effectiveness.

Information-Theoretical Principled Trade-off between Jailbreakability and Stealthiness on Vision Language Models

In recent years, Vision-Language Models (VLMs) have demonstrated significant advancements in artificial intelligence, transforming tasks across various domains. Despite their capabilities, these models are susceptible to jailbreak attacks, which can compromise their safety and reliability. This paper explores the trade-off between jailbreakability and stealthiness in VLMs, presenting a novel algorithm to detect non-stealthy jailbreak attacks and enhance model robustness. We introduce a stealthiness-aware jailbreak attack using diffusion models, highlighting the challenge of detecting AI-generated content. Our approach leverages Fano's inequality to elucidate the relationship between attack success rates and stealthiness scores, providing an explainable framework for evaluating these threats. Our contributions aim to fortify AI systems against sophisticated attacks, ensuring their outputs remain aligned with ethical standards and user expectations.

Exploring Robustness of Visual State Space model against Backdoor Attacks

Visual State Space Model (VSS) has demonstrated remarkable performance in various computer vision tasks. However, in the process of development, backdoor attacks have brought severe challenges to security. Such attacks cause an infected model to predict target labels when a specific trigger is activated, while the model behaves normally on benign samples. In this paper, we conduct systematic experiments to comprehend on robustness of VSS through the lens of backdoor attacks, specifically how the state space model (SSM) mechanism affects robustness. We first investigate the vulnerability of VSS to different backdoor triggers and reveal that the SSM mechanism, which captures contextual information within patches, makes the VSS model more susceptible to backdoor triggers compared to models without SSM. Furthermore, we analyze the sensitivity of the VSS model to patch processing techniques and discover that these triggers are effectively disrupted. Based on these observations, we consider an effective backdoor for the VSS model that recurs in each patch to resist patch perturbations. Extensive experiments across three datasets and various backdoor attacks reveal that the VSS model performs comparably to Transformers (ViTs) but is less robust than the Gated CNNs, which comprise only stacked Gated CNN blocks without SSM.

Defending Against Repetitive-based Backdoor Attacks on Semi-supervised Learning through Lens of Rate-Distortion-Perception Trade-off

Semi-supervised learning (SSL) has achieved remarkable performance with a small fraction of labeled data by leveraging vast amounts of unlabeled data from the Internet. However, this large pool of untrusted data is extremely vulnerable to data poisoning, leading to potential backdoor attacks. Current backdoor defenses are not yet effective against such a vulnerability in SSL. In this study, we propose a novel method, Unlabeled Data Purification (UPure), to disrupt the association between trigger patterns and target classes by introducing perturbations in the frequency domain. By leveraging the Rate- Distortion-Perception (RDP) trade-off, we further identify the frequency band, where the perturbations are added, and justify this selection. Notably, UPure purifies poisoned unlabeled data without the need of extra clean labeled data. Extensive experiments on four benchmark datasets and five SSL algorithms demonstrate that UPure effectively reduces the attack success rate from 99.78% to 0% while maintaining model accuracy

Safe LoRA: the Silver Lining of Reducing Safety Risks when Fine-tuning Large Language Models

While large language models (LLMs) such as Llama-2 or GPT-4 have shown impressive zero-shot performance, fine-tuning is still necessary to enhance their performance for customized datasets, domain-specific tasks, or other private needs. However, fine-tuning all parameters of LLMs requires significant hardware resources, which can be impractical for typical users. Therefore, parameter-efficient fine-tuning such as LoRA have emerged, allowing users to fine-tune LLMs without the need for considerable computing resources, with little performance degradation compared to fine-tuning all parameters. Unfortunately, recent studies indicate that fine-tuning can increase the risk to the safety of LLMs, even when data does not contain malicious content. To address this challenge, we propose Safe LoRA, a simple one-liner patch to the original LoRA implementation by introducing the projection of LoRA weights from selected layers to the safety-aligned subspace, effectively reducing the safety risks in LLM fine-tuning while maintaining utility. It is worth noting that Safe LoRA is a training-free and data-free approach, as it only requires the knowledge of the weights from the base and aligned LLMs. Our extensive experiments demonstrate that when fine-tuning on purely malicious data, Safe LoRA retains similar safety performance as the original aligned model. Moreover, when the fine-tuning dataset contains a mixture of both benign and malicious data, Safe LoRA mitigates the negative effect made by malicious data while preserving performance on downstream tasks.

DiffuseKronA: A Parameter Efficient Fine-tuning Method for Personalized Diffusion Models

In the realm of subject-driven text-to-image (T2I) generative models, recent developments like DreamBooth and BLIP-Diffusion have led to impressive results yet encounter limitations due to their intensive fine-tuning demands and substantial parameter requirements. While the low-rank adaptation (LoRA) module within DreamBooth offers a reduction in trainable parameters, it introduces a pronounced sensitivity to hyperparameters, leading to a compromise between parameter efficiency and the quality of T2I personalized image synthesis. Addressing these constraints, we introduce \textbf{\textit{DiffuseKronA}}, a novel Kronecker product-based adaptation module that not only significantly reduces the parameter count by 35\% and 99.947\% compared to LoRA-DreamBooth and the original DreamBooth, respectively, but also enhances the quality of image synthesis. Crucially, \textit{DiffuseKronA} mitigates the issue of hyperparameter sensitivity, delivering consistent high-quality generations across a wide range of hyperparameters, thereby diminishing the necessity for extensive fine-tuning. Furthermore, a more controllable decomposition makes \textit{DiffuseKronA} more interpretable and even can achieve up to a 50\% reduction with results comparable to LoRA-Dreambooth. Evaluated against diverse and complex input images and text prompts, \textit{DiffuseKronA} consistently outperforms existing models, producing diverse images of higher quality with improved fidelity and a more accurate color distribution of objects, all the while upholding exceptional parameter efficiency, thus presenting a substantial advancement in the field of T2I generative modeling. Our project page, consisting of links to the code, and pre-trained checkpoints, is available at https://diffusekrona.github.io/.

Rethinking Backdoor Attacks on Dataset Distillation: A Kernel Method Perspective

Dataset distillation offers a potential means to enhance data efficiency in deep learning. Recent studies have shown its ability to counteract backdoor risks present in original training samples. In this study, we delve into the theoretical aspects of backdoor attacks and dataset distillation based on kernel methods. We introduce two new theory-driven trigger pattern generation methods specialized for dataset distillation. Following a comprehensive set of analyses and experiments, we show that our optimization-based trigger design framework informs effective backdoor attacks on dataset distillation. Notably, datasets poisoned by our designed trigger prove resilient against conventional backdoor attack detection and mitigation methods. Our empirical results validate that the triggers developed using our approaches are proficient at executing resilient backdoor attacks.

Ring-A-Bell! How Reliable are Concept Removal Methods for Diffusion Models?

Diffusion models for text-to-image (T2I) synthesis, such as Stable Diffusion (SD), have recently demonstrated exceptional capabilities for generating high-quality content. However, this progress has raised several concerns of potential misuse, particularly in creating copyrighted, prohibited, and restricted content, or NSFW (not safe for work) images. While efforts have been made to mitigate such problems, either by implementing a safety filter at the evaluation stage or by fine-tuning models to eliminate undesirable concepts or styles, the effectiveness of these safety measures in dealing with a wide range of prompts remains largely unexplored. In this work, we aim to investigate these safety mechanisms by proposing one novel concept retrieval algorithm for evaluation. We introduce Ring-A-Bell, a model-agnostic red-teaming tool for T2I diffusion models, where the whole evaluation can be prepared in advance without prior knowledge of the target model. Specifically, Ring-A-Bell first performs concept extraction to obtain holistic representations for sensitive and inappropriate concepts. Subsequently, by leveraging the extracted concept, Ring-A-Bell automatically identifies problematic prompts for diffusion models with the corresponding generation of inappropriate content, allowing the user to assess the reliability of deployed safety mechanisms. Finally, we empirically validate our method by testing online services such as Midjourney and various methods of concept removal. Our results show that Ring-A-Bell, by manipulating safe prompting benchmarks, can transform prompts that were originally regarded as safe to evade existing safety mechanisms, thus revealing the defects of the so-called safety mechanisms which could practically lead to the generation of harmful contents.

Exploring the Benefits of Differentially Private Pre-training and Parameter-Efficient Fine-tuning for Table Transformers

For machine learning with tabular data, Table Transformer (TabTransformer) is a state-of-the-art neural network model, while Differential Privacy (DP) is an essential component to ensure data privacy. In this paper, we explore the benefits of combining these two aspects together in the scenario of transfer learning -- differentially private pre-training and fine-tuning of TabTransformers with a variety of parameter-efficient fine-tuning (PEFT) methods, including Adapter, LoRA, and Prompt Tuning. Our extensive experiments on the ACSIncome dataset show that these PEFT methods outperform traditional approaches in terms of the accuracy of the downstream task and the number of trainable parameters, thus achieving an improved trade-off among parameter efficiency, privacy, and accuracy. Our code is available at github.com/IBM/DP-TabTransformer.

DPAF: Image Synthesis via Differentially Private Aggregation in Forward Phase

Differentially private synthetic data is a promising alternative for sensitive data release. Many differentially private generative models have been proposed in the literature. Unfortunately, they all suffer from the low utility of the synthetic data, particularly for images of high resolutions. Here, we propose DPAF, an effective differentially private generative model for high-dimensional image synthesis. Different from the prior private stochastic gradient descent-based methods that add Gaussian noises in the backward phase during the model training, DPAF adds a differentially private feature aggregation in the forward phase, bringing advantages, including the reduction of information loss in gradient clipping and low sensitivity for the aggregation. Moreover, as an improper batch size has an adverse impact on the utility of synthetic data, DPAF also tackles the problem of setting a proper batch size by proposing a novel training strategy that asymmetrically trains different parts of the discriminator. We extensively evaluate different methods on multiple image datasets (up to images of 128x128 resolution) to demonstrate the performance of DPAF.