The emergence of large language models (LLMs), and their increased use in user-facing systems, has led to substantial privacy concerns. To date, research on these privacy concerns has been model-centered: exploring how LLMs lead to privacy risks like memorization, or can be used to infer personal characteristics about people from their content. We argue that there is a need for more research focusing on the human aspect of these privacy issues: e.g., research on how design paradigms for LLMs affect users' disclosure behaviors, users' mental models and preferences for privacy controls, and the design of tools, systems, and artifacts that empower end-users to reclaim ownership over their personal data. To build usable, efficient, and privacy-friendly systems powered by these models with imperfect privacy properties, our goal is to initiate discussions to outline an agenda for conducting human-centered research on privacy issues in LLM-powered systems. This Special Interest Group (SIG) aims to bring together researchers with backgrounds in usable security and privacy, human-AI collaboration, NLP, or any other related domains to share their perspectives and experiences on this problem, to help our community establish a collective understanding of the challenges, research opportunities, research methods, and strategies to collaborate with researchers outside of HCI.
相關內容
Spectrum-Based Fault Localization (SBFL) is a technique to be used during debugging, the premise of which is that, based on the test case outcomes and code coverage, faulty code elements can be automatically detected. SBFL is popular among researchers because it is lightweight and easy to implement, and there is a lot of potential in it when it comes to research that aims to improve its effectiveness. Despite this, the technique cannot be found in contemporary development and debugging tools, only a handful of research prototypes are available. Reasons for this can be multiple, including the algortihms' sub-optimal effectiveness and other technical weaknesses. But, also the lack of clear functional and non-functional requirements for such a tool, either standalone or integrated into IDEs. In this paper, we attempt to provide such a list in form of recommendations, based on surveying the most popular SBFL tools and on our own researchers' and tool builders' experience.
The explanations of large language models have recently been shown to be sensitive to the randomness used for their training, creating a need to characterize this sensitivity. In this paper, we propose a characterization that questions the possibility to provide simple and informative explanations for such models. To this end, we give statistical definitions for the explanations' signal, noise and signal-to-noise ratio. We highlight that, in a typical case study where word-level univariate explanations are analyzed with first-order statistical tools, the explanations of simple feature-based models carry more signal and less noise than those of transformer ones. We then discuss the possibility to improve these results with alternative definitions of signal and noise that would capture more complex explanations and analysis methods, while also questioning the tradeoff with their plausibility for readers.
The presence of toxic and gender-identity derogatory language in open-source software (OSS) communities has recently become a focal point for researchers. Such comments not only lead to frustration and disengagement among developers but may also influence their leave from the OSS projects. Despite ample evidence suggesting that diverse teams enhance productivity, the existence of toxic or gender identity discriminatory communications poses a significant threat to the participation of individuals from marginalized groups and, as such, may act as a barrier to fostering diversity and inclusion in OSS projects. However, there is a notable lack of research dedicated to exploring the association between gender-based toxic and derogatory language with a perceptible diversity of open-source software teams. Consequently, this study aims to investigate how such content influences the gender, ethnicity, and tenure diversity of open-source software development teams. To achieve this, we extract data from active GitHub projects, assess various project characteristics, and identify instances of toxic and gender-discriminatory language within issue/pull request comments. Using these attributes, we construct a regression model to explore how they associate with the perceptible diversity of those projects.
Generative text-to-image models, which allow users to create appealing images through a text prompt, have seen a dramatic increase in popularity in recent years. However, most users have a limited understanding of how such models work and it often requires many trials and errors to achieve satisfactory results. The prompt history contains a wealth of information that could provide users with insights into what have been explored and how the prompt changes impact the output image, yet little research attention has been paid to the visual analysis of such process to support users. We propose the Image Variant Graph, a novel visual representation designed to support comparing prompt-image pairs and exploring the editing history. The Image Variant Graph models prompt differences as edges between corresponding images and presents the distances between images through projection. Based on the graph, we developed the PrompTHis system through co-design with artists. Besides Image Variant Graph, PrompTHis also incorporates a detailed prompt-image history and a navigation mini-map. Based on the review and analysis of the prompting history, users can better understand the impact of prompt changes and have a more effective control of image generation. A quantitative user study with eleven amateur participants and qualitative interviews with five professionals and one amateur user were conducted to evaluate the effectiveness of PrompTHis. The results demonstrate PrompTHis can help users review the prompt history, make sense of the model, and plan their creative process.
As machine learning models become increasingly larger, trained weakly supervised on large, possibly uncurated data sets, it becomes increasingly important to establish mechanisms for inspecting, interacting, and revising models to mitigate learning shortcuts and guarantee their learned knowledge is aligned with human knowledge. The recently proposed XIL framework was developed for this purpose, and several such methods have been introduced, each with individual motivations and methodological details. In this work, we provide a unification of various XIL methods into a single typology by establishing a common set of basic modules. In doing so, we pave the way for a principled comparison of existing, but, importantly, also future XIL approaches. In addition, we discuss existing and introduce novel measures and benchmarks for evaluating the overall abilities of a XIL method. Given this extensive toolbox, including our typology, measures, and benchmarks, we finally compare several recent XIL methods methodologically and quantitatively. In our evaluations, all methods prove to revise a model successfully. However, we found remarkable differences in individual benchmark tasks, revealing valuable application-relevant aspects for integrating these benchmarks in developing future methods.
Pretrained language models are commonly aligned with human preferences and downstream tasks via reinforcement finetuning (RFT), which refers to maximizing a (possibly learned) reward function using policy gradient algorithms. This work identifies a fundamental optimization obstacle in RFT: we prove that the expected gradient for an input vanishes when its reward standard deviation under the model is small, even if the expected reward is far from optimal. Through experiments on an RFT benchmark and controlled environments, as well as a theoretical analysis, we then demonstrate that vanishing gradients due to small reward standard deviation are prevalent and detrimental, leading to extremely slow reward maximization. Lastly, we explore ways to overcome vanishing gradients in RFT. We find the common practice of an initial supervised finetuning (SFT) phase to be the most promising candidate, which sheds light on its importance in an RFT pipeline. Moreover, we show that a relatively small number of SFT optimization steps on as few as 1% of the input samples can suffice, indicating that the initial SFT phase need not be expensive in terms of compute and data labeling efforts. Overall, our results emphasize that being mindful for inputs whose expected gradient vanishes, as measured by the reward standard deviation, is crucial for successful execution of RFT.
Natural language processing models have experienced a significant upsurge in recent years, with numerous applications being built upon them. Many of these applications require fine-tuning generic base models on customized, proprietary datasets. This fine-tuning data is especially likely to contain personal or sensitive information about individuals, resulting in increased privacy risk. Membership inference attacks are the most commonly employed attack to assess the privacy leakage of a machine learning model. However, limited research is available on the factors that affect the vulnerability of language models to this kind of attack, or on the applicability of different defense strategies in the language domain. We provide the first systematic review of the vulnerability of fine-tuned large language models to membership inference attacks, the various factors that come into play, and the effectiveness of different defense strategies. We find that some training methods provide significantly reduced privacy risk, with the combination of differential privacy and low-rank adaptors achieving the best privacy protection against these attacks.
While large language models (LLMs) have demonstrated remarkable capabilities across a range of downstream tasks, a significant concern revolves around their propensity to exhibit hallucinations: LLMs occasionally generate content that diverges from the user input, contradicts previously generated context, or misaligns with established world knowledge. This phenomenon poses a substantial challenge to the reliability of LLMs in real-world scenarios. In this paper, we survey recent efforts on the detection, explanation, and mitigation of hallucination, with an emphasis on the unique challenges posed by LLMs. We present taxonomies of the LLM hallucination phenomena and evaluation benchmarks, analyze existing approaches aiming at mitigating LLM hallucination, and discuss potential directions for future research.
In pace with developments in the research field of artificial intelligence, knowledge graphs (KGs) have attracted a surge of interest from both academia and industry. As a representation of semantic relations between entities, KGs have proven to be particularly relevant for natural language processing (NLP), experiencing a rapid spread and wide adoption within recent years. Given the increasing amount of research work in this area, several KG-related approaches have been surveyed in the NLP research community. However, a comprehensive study that categorizes established topics and reviews the maturity of individual research streams remains absent to this day. Contributing to closing this gap, we systematically analyzed 507 papers from the literature on KGs in NLP. Our survey encompasses a multifaceted review of tasks, research types, and contributions. As a result, we present a structured overview of the research landscape, provide a taxonomy of tasks, summarize our findings, and highlight directions for future work.
Deep neural networks have revolutionized many machine learning tasks in power systems, ranging from pattern recognition to signal processing. The data in these tasks is typically represented in Euclidean domains. Nevertheless, there is an increasing number of applications in power systems, where data are collected from non-Euclidean domains and represented as the graph-structured data with high dimensional features and interdependency among nodes. The complexity of graph-structured data has brought significant challenges to the existing deep neural networks defined in Euclidean domains. Recently, many studies on extending deep neural networks for graph-structured data in power systems have emerged. In this paper, a comprehensive overview of graph neural networks (GNNs) in power systems is proposed. Specifically, several classical paradigms of GNNs structures (e.g., graph convolutional networks, graph recurrent neural networks, graph attention networks, graph generative networks, spatial-temporal graph convolutional networks, and hybrid forms of GNNs) are summarized, and key applications in power systems such as fault diagnosis, power prediction, power flow calculation, and data generation are reviewed in detail. Furthermore, main issues and some research trends about the applications of GNNs in power systems are discussed.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191