Related papers: Easy to Decide, Hard to Agree: Reducing Disagreeme…

Deep Learning of neural networks has progressively become more prominent in healthcare with models reaching, or even surpassing, expert accuracy levels. However, these success stories are tainted by concerning reports on the lack of model…

Probabilistic confidence metrics are increasingly adopted as proxies for reasoning quality in Best-of-N selection, under the assumption that higher confidence reflects higher reasoning fidelity. In this work, we challenge this assumption by…

In this thesis, I refine our understanding as to what conclusions we can reach from coreference-based evaluations by expanding existing evaluation practices and considering the extent to which evaluation results are either converging or…

We examine diverging preferences in human-labeled preference datasets. We develop a taxonomy of disagreement sources spanning ten categories across four high-level classes and find that the majority of disagreements are due to factors such…

Recently, state-of-the-art NLP models gained an increasing syntactic and semantic understanding of language, and explanation methods are crucial to understand their decisions. Occlusion is a well established method that provides…

We describe a context-sensitive model of choice, in which the selection process is shaped not only by the attractiveness of items but also by their semantics ('salience'). All items are ranked according to a relation of salience, and a…

First derived from human intuition, later adapted to machine translation for automatic token alignment, attention mechanism, a simple method that can be used for encoding sequence data based on the importance score each element is assigned,…

While deep neural networks have achieved remarkable performance, they tend to lack transparency in prediction. The pursuit of greater interpretability in neural networks often results in a degradation of their original performance. Some…

With the ever-increasing complexity of neural language models, practitioners have turned to methods for understanding the predictions of these models. One of the most well-adopted approaches for model interpretability is feature-based…

In the field of Explainable AI, multiples evaluation metrics have been proposed in order to assess the quality of explanation methods w.r.t. a set of desired properties. In this work, we study the articulation between the stability,…

Weighted knowledge bases for description logics with typicality under a "concept-wise" multi-preferential semantics provide a logical interpretation of MultiLayer Perceptrons. In this context, Answer Set Programming (ASP) has been shown to…

Variation in human annotation (i.e., disagreements) is common in NLP, often reflecting important information like task subjectivity and sample ambiguity. Modeling this variation is important for applications that are sensitive to such…

Clinical applicability of automated decision support systems depends on a robust, well-understood classification interpretation. Artificial neural networks while achieving class-leading scores fall short in this regard. Therefore, numerous…

Recent studies on the adversarial vulnerability of neural networks have shown that models trained to be more robust to adversarial attacks exhibit more interpretable saliency maps than their non-robust counterparts. We aim to quantify this…

A distinction is often drawn between a model's ability to predict a label for an evaluation sample that is directly memorised from highly similar training samples versus an ability to predict the label via some method of generalisation. In…

Methods based on class activation maps (CAM) provide a simple mechanism to interpret predictions of convolutional neural networks by using linear combinations of feature maps as saliency maps. By contrast, masking-based methods optimize a…

Existing approaches to explaining deep learning models in NLP usually suffer from two major drawbacks: (1) the main model and the explaining model are decoupled: an additional probing or surrogate model is used to interpret an existing…

Saliency methods have been widely used to highlight important input features in model predictions. Most existing methods use backpropagation on a modified gradient function to generate saliency maps. Thus, noisy gradients can result in…

Understanding why a classification model prefers one class over another for an input instance is the challenge of contrastive explanation. This work implements concept-based contrastive explanations for image classification by leveraging…

Attention mechanisms play a central role in NLP systems, especially within recurrent neural network (RNN) models. Recently, there has been increasing interest in whether or not the intermediate representations offered by these modules may…