This paper introduces a novel explainable artificial intelligence (XAI) approach based on near-misses analysis (NMA). This approach uses the network close related predictions to reveal a hierarchy of logical concepts inferred from the latent decision-making process of a neural network (NN) without delving into its explicit structure. Several NMA usage possibilities are reported in this paper. First, it serves to create an explanation in the form of a gradually expanding explicit linked concepts which coupled with a proper dictionary can provide a scoring method to differentiate which of any given models is better at providing human-like conceptual explanations. In addition, NMA can be used to pinpoint how to improve models according to their explanatory outcome. Finally, it enables to detect adversarial attacks. The proposed XAI approach is examined on different network architectures that vary in size and shape (ResNet, VGG, EfficientNet, MobileNet) and on several datasets which were already organized in a hierarchical concept structure (ImageNet, CIFAR100). Results demonstrate that NMA can reflect NNs latent concepts generation process. Moreover, using the devised scoring method, it is reported that efficient architectures, which achieve a similar accuracy level with less neurons, may still pay the price of explainability and robustness. • Introduce a new XAI method based on neural networks concepts generation. • Demonstrate the explanatory capabilities of this method on various models and datasets. • Propose a new explanatory scoring method. • Demonstrate how to use the new method for model debugging. • Demonstrate how to use the new method for detecting adversarial attacks.

Even with various regulations in place across countries and social media platforms (Government of India, 2021; European Parliament and Council of the European Union, 2022), digital abusive speech remains a significant issue. One potential approach to address this challenge is automatic text detoxification, a text style transfer (TST) approach that transforms toxic language into a more neutral or non-toxic form. To date, the availability of parallel corpora for the text detoxification task (Logacheva et al., 2022; Atwell et al., 2022; Dementieva et al., 2024a) has proven to be crucial for state-of-the-art approaches. With this work, we extend parallel text detoxification corpus to new languages—German, Chinese, Arabic, Hindi, and Amharic—testing in the extensive multilingual setup TST baselines. Next, we conduct the first of its kind an automated, explainable analysis of the descriptive features of both toxic and non-toxic sentences, diving deeply into the nuances, similarities, and differences of toxicity and detoxification across 9 languages. Finally, based on the obtained insights, we experiment with a novel text detoxification method inspired by the Chain-of-Thoughts reasoning approach, enhancing the prompting process through clustering on relevant descriptive attributes.

Abstract We introduce a new embedding technique based on a nested barycentric coordinate system. We show that our embedding can be used to transform the problems of polyhedron approximation, piecewise linear classification and convex regression into one of finding a linear classifier or regressor in a higher dimensional (but nevertheless quite sparse) representation. Our embedding maps a piecewise linear function into an everywhere-linear function, and allows us to invoke well-known algorithms for the latter problem to solve the former. We explain the applications of our embedding to the problems of approximating separating polyhedra—in fact, it can approximate any convex body and unions of convex bodies—as well as to classification by separating polyhedra, and to piecewise linear regression.

We present an improved algorithm forquasi-properlylearning convex polyhedra in the realizable PAC setting from data with a margin. Our learning algorithm constructs a consistent polyhedron as an intersection of about$t \log t$halfspaces with constant-size margins in time polynomial in$t$(where$t$is the number of halfspaces forming an optimal polyhedron). We also identify distinct generalizations of the notion of margin from hyperplanes to polyhedra and investigate how they relate geometrically; this result may have ramifications beyond the learning setting.

We introduce a new embedding technique based on barycentric coordinate system. We show that our embedding can be used to transforms the problem of polytope approximation into that of finding a linear classifier in a higher (but nevertheless quite sparse) dimensional representation. This embedding in effect maps a piecewise linear function into a single linear function, and allows us to invoke well-known algorithms for the latter problem to solve the former. We demonstrate that our embedding has applications to the problems of approximating separating polytopes – in fact, it can approximate any convex body and multiple convex bodies – as well as to classification by separating polytopes and piecewise linear regression.

We present improved algorithm for properly learning convex polytopes in the realizable PAC setting from data with a margin. Our learning algorithm constructs a consistent polytope as an intersection of about t log t halfspaces with margins in time polynomial in t (where t is the number of halfspaces forming an optimal polytope). We also identify distinct generalizations of the notion of margin from hyperplanes to polytopes and investigate how they relate geometrically; this result may be of interest beyond the learning setting.