JailbreakLens: Visual Analysis of Jailbreak Attacks Against Large Language Models

The proliferation of large language models (LLMs) has underscored concerns regarding their security vulnerabilities, notably against jailbreak attacks, where adversaries design jailbreak prompts to circumvent safety mechanisms for potential misuse. Addressing these concerns necessitates a comprehensive analysis of jailbreak prompts to evaluate LLMs' defensive capabilities and identify potential weaknesses. However, the complexity of evaluating jailbreak performance and understanding prompt characteristics makes this analysis laborious. We collaborate with domain experts to characterize problems and propose an LLM-assisted framework to streamline the analysis process. It provides automatic jailbreak assessment to facilitate performance evaluation and support analysis of components and keywords in prompts. Based on the framework, we design JailbreakLens, a visual analysis system that enables users to explore the jailbreak performance against the target model, conduct multi-level analysis of prompt characteristics, and refine prompt instances to verify findings. Through a case study, technical evaluations, and expert interviews, we demonstrate our system's effectiveness in helping users evaluate model security and identify model weaknesses.

Self-Distillation Bridges Distribution Gap in Language Model Fine-Tuning

The surge in Large Language Models (LLMs) has revolutionized natural language processing, but fine-tuning them for specific tasks often encounters challenges in balancing performance and preserving general instruction-following abilities. In this paper, we posit that the distribution gap between task datasets and the LLMs serves as the primary underlying cause. To address the problem, we introduce Self-Distillation Fine-Tuning (SDFT), a novel approach that bridges the distribution gap by guiding fine-tuning with a distilled dataset generated by the model itself to match its original distribution. Experimental results on the Llama-2-chat model across various benchmarks demonstrate that SDFT effectively mitigates catastrophic forgetting while achieving comparable or superior performance on downstream tasks compared to the vanilla fine-tuning. Moreover, SDFT demonstrates the potential to maintain the helpfulness and safety alignment of LLMs. Our code is available at \url{https://github.com/sail-sg/sdft}.

PromptMagician: Interactive Prompt Engineering for Text-to-Image Creation

Generative text-to-image models have gained great popularity among the public for their powerful capability to generate high-quality images based on natural language prompts. However, developing effective prompts for desired images can be challenging due to the complexity and ambiguity of natural language. This research proposes PromptMagician, a visual analysis system that helps users explore the image results and refine the input prompts. The backbone of our system is a prompt recommendation model that takes user prompts as input, retrieves similar prompt-image pairs from DiffusionDB, and identifies special (important and relevant) prompt keywords. To facilitate interactive prompt refinement, PromptMagician introduces a multi-level visualization for the cross-modal embedding of the retrieved images and recommended keywords, and supports users in specifying multiple criteria for personalized exploration. Two usage scenarios, a user study, and expert interviews demonstrate the effectiveness and usability of our system, suggesting it facilitates prompt engineering and improves the creativity support of the generative text-to-image model.

CoSDA: Continual Source-Free Domain Adaptation

Without access to the source data, source-free domain adaptation (SFDA) transfers knowledge from a source-domain trained model to target domains. Recently, SFDA has gained popularity due to the need to protect the data privacy of the source domain, but it suffers from catastrophic forgetting on the source domain due to the lack of data. To systematically investigate the mechanism of catastrophic forgetting, we first reimplement previous SFDA approaches within a unified framework and evaluate them on four benchmarks. We observe that there is a trade-off between adaptation gain and forgetting loss, which motivates us to design a consistency regularization to mitigate forgetting. In particular, we propose a continual source-free domain adaptation approach named CoSDA, which employs a dual-speed optimized teacher-student model pair and is equipped with consistency learning capability. Our experiments demonstrate that CoSDA outperforms state-of-the-art approaches in continuous adaptation. Notably, our CoSDA can also be integrated with other SFDA methods to alleviate forgetting.

Interactive Image Synthesis with Panoptic Layout Generation

Interactive image synthesis from user-guided input is a challenging task when users wish to control the scene structure of a generated image with ease.Although remarkable progress has been made on layout-based image synthesis approaches, in order to get realistic fake image in interactive scene, existing methods require high-precision inputs, which probably need adjustment several times and are unfriendly to novice users. When placement of bounding boxes is subject to perturbation, layout-based models suffer from "missing regions" in the constructed semantic layouts and hence undesirable artifacts in the generated images. In this work, we propose Panoptic Layout Generative Adversarial Networks (PLGAN) to address this challenge. The PLGAN employs panoptic theory which distinguishes object categories between "stuff" with amorphous boundaries and "things" with well-defined shapes, such that stuff and instance layouts are constructed through separate branches and later fused into panoptic layouts. In particular, the stuff layouts can take amorphous shapes and fill up the missing regions left out by the instance layouts. We experimentally compare our PLGAN with state-of-the-art layout-based models on the COCO-Stuff, Visual Genome, and Landscape datasets. The advantages of PLGAN are not only visually demonstrated but quantitatively verified in terms of inception score, Fr\'echet inception distance, classification accuracy score, and coverage.

KD3A: Unsupervised Multi-Source Decentralized Domain Adaptation via Knowledge Distillation

Conventional unsupervised multi-source domain adaptation (UMDA) methods assume all source domains can be accessed directly. This neglects the privacy-preserving policy, that is, all the data and computations must be kept decentralized. There exists three problems in this scenario: (1) Minimizing the domain distance requires the pairwise calculation of the data from source and target domains, which is not accessible. (2) The communication cost and privacy security limit the application of UMDA methods (e.g., the domain adversarial training). (3) Since users have no authority to check the data quality, the irrelevant or malicious source domains are more likely to appear, which causes negative transfer. In this study, we propose a privacy-preserving UMDA paradigm named Knowledge Distillation based Decentralized Domain Adaptation (KD3A), which performs domain adaptation through the knowledge distillation on models from different source domains. KD3A solves the above problems with three components: (1) A multi-source knowledge distillation method named Knowledge Vote to learn high-quality domain consensus knowledge. (2) A dynamic weighting strategy named Consensus Focus to identify both the malicious and irrelevant domains. (3) A decentralized optimization strategy for domain distance named BatchNorm MMD. The extensive experiments on DomainNet demonstrate that KD3A is robust to the negative transfer and brings a 100x reduction of communication cost compared with other decentralized UMDA methods. Moreover, our KD3A significantly outperforms state-of-the-art UMDA approaches.

SHOT-VAE: Semi-supervised Deep Generative Models With Label-aware ELBO Approximations

Semi-supervised variational autoencoders (VAEs) have obtained strong results, but have also encountered the challenge that good ELBO values do not always imply accurate inference results. In this paper, we investigate and propose two causes of this problem: (1) The ELBO objective cannot utilize the label information directly. (2) A bottleneck value exists and continuing to optimize ELBO after this value will not improve inference accuracy. On the basis of the experiment results, we propose SHOT-VAE to address these problems without introducing additional prior knowledge. The SHOT-VAE offers two contributions: (1) A new ELBO approximation named smooth-ELBO that integrates the label predictive loss into ELBO. (2) An approximation based on optimal interpolation that breaks the ELBO value bottleneck by reducing the margin between ELBO and the data likelihood. The SHOT-VAE achieves good performance with a 25.30% error rate on CIFAR-100 with 10k labels and reduces the error rate to 6.11% on CIFAR-10 with 4k labels.

DM-GAN: Dynamic Memory Generative Adversarial Networks for Text-to-Image Synthesis

In this paper, we focus on generating realistic images from text descriptions. Current methods first generate an initial image with rough shape and color, and then refine the initial image to a high-resolution one. Most existing text-to-image synthesis methods have two main problems. (1) These methods depend heavily on the quality of the initial images. If the initial image is not well initialized, the following processes can hardly refine the image to a satisfactory quality. (2) Each word contributes a different level of importance when depicting different image contents, however, unchanged text representation is used in existing image refinement processes. In this paper, we propose the Dynamic Memory Generative Adversarial Network (DM-GAN) to generate high-quality images. The proposed method introduces a dynamic memory module to refine fuzzy image contents, when the initial images are not well generated. A memory writing gate is designed to select the important text information based on the initial image content, which enables our method to accurately generate images from the text description. We also utilize a response gate to adaptively fuse the information read from the memories and the image features. We evaluate the DM-GAN model on the Caltech-UCSD Birds 200 dataset and the Microsoft Common Objects in Context dataset. Experimental results demonstrate that our DM-GAN model performs favorably against the state-of-the-art approaches.