Large Language Models (LLMs) have been shown to be susceptible to jailbreak attacks, or adversarial attacks used to illicit high risk behavior from a model, highlighting the critical need to safeguard widely-deployed models.Safeguarding approaches, which include finetuning models or having LLMs "self-reflect," may lengthen the inference time of a model, incur a computational penalty, reduce the semantic fluency of an output, and restrict "normal" model behavior.Importantly, these Safety-Performance Trade-offs (SPTs) remain an understudied area.In this work, we make three contributions: (1) We introduce SAFENUDGE, a novel safeguard that combines Controlled Text Generation and "nudging."SAFENUDGE triggers during text-generation while a jailbreak attack is being executed, and can reduce successful jailbreak attempts by between 28.1% and 37.3% by guiding the LLM towards a safe response.It adds minimal latency to inference and has a negligible impact on the semantic fluency of outputs.Second, it supports tunable SPTs, meaning practitioners can set their own tolerance for tradeoffs balancing safety and restrictions to normal model behavior.Third, we release the source code for SAFENUDGE at https:// github.com/joaopfonseca/SafeNudge.It is open source and compatible with the Hugging Face transformers library.

Algorithmic decisions in critical domains such as hiring, college admissions, and lending are often based on rankings. Given the impact of these decisions on individuals, organizations, and population groups, it is essential to understand them—to help individuals improve their ranking position, design better ranking procedures, and ensure legal compliance. In this paper, we argue that explainability methods for classification and regression, such as SHAP, are insufficient for ranking tasks, and present ShaRP—Shapley Values for Rankings and Preferences—a framework that explains the contributions of features to various aspects of a ranked outcome. ShaRP computes feature contributions for various ranking-specific profit functions, such as rank and top- k , and also includes a novel Shapley value-based method for explaining pairwise preference outcomes. We provide a flexible implementation of ShaRP, capable of efficiently and comprehensively explaining ranked and pairwise outcomes over tabular data, in score-based ranking and learning-to-rank tasks. Finally, we develop a comprehensive evaluation methodology for ranking explainability methods, showing through qualitative, quantitative, and usability studies that our rank-aware QoIs offer complementary insights, scale effectively, and help users interpret ranked outcomes in practice.

Imbalanced learning can be addressed in 3 different ways: Resampling, algorithmic modifications and cost-sensitive solutions. Resampling, and specifically oversampling, are more general approaches when opposed to algorithmic and cost-sensitive methods. Since the proposal of the Synthetic Minority Oversampling TEchnique (SMOTE), various SMOTE variants and neural network-based oversampling methods have been developed. However, the options to oversample datasets with nominal and continuous features are limited. We propose Geometric SMOTE for Nominal and Continuous features (G-SMOTENC), based on a combination of G-SMOTE and SMOTENC. Our method modifies SMOTENC’s encoding and generation mechanism for nominal features while using G-SMOTE’s data selection mechanism to determine the center observation and k-nearest neighbors and generation mechanism for continuous features. G-SMOTENC’s performance is compared against SMOTENC’s along with two other baseline methods, a State-of-the-art oversampling method and no oversampling. The experiment was performed over 20 datasets with varying imbalance ratios, number of metric and non-metric features and target classes. We found a significant improvement in classification performance when using G-SMOTENC as the oversampling method. An open-source implementation of G-SMOTENC is made available in the Python programming language.

Active learning (AL) is a well‐known technique to optimize data usage in training, through the interactive selection of unlabeled observations, out of a large pool of unlabeled data, to be labeled by a supervisor. Its focus is to find the unlabeled observations that, once labeled, will maximize the informativeness of the training dataset, therefore reducing data‐related costs. The literature describes several methods to improve the effectiveness of this process. Nonetheless, there is a paucity of research developed around the application of artificial data sources in AL, especially outside image classification or NLP. This paper proposes a new AL framework, which relies on the effective use of artificial data. It may be used with any classifier, generation mechanism, and data type and can be integrated with multiple other state‐of‐the‐art AL contributions. This combination is expected to increase the ML classifier’s performance and reduce both the supervisor’s involvement and the amount of required labeled data at the expense of a marginal increase in computational time. The proposed method introduces a hyperparameter optimization component to improve the generation of artificial instances during the AL process as well as an uncertainty‐based data generation mechanism. We compare the proposed method to the standard framework and an oversampling‐based active learning method for more informed data generation in an AL context. The models’ performance was tested using four different classifiers, two AL‐specific performance metrics, and three classification performance metrics over 15 different datasets. We demonstrated that the proposed framework, using data augmentation, significantly improved the performance of AL, both in terms of classification performance and data selection efficiency (all the codes and preprocessed data developed for this study are available at https://github.com/joaopfonseca/publications/ ).