Research Papers

The evolution of Explainable Artificial Intelligence (XAI) has emphasised the significance of meeting diverse user needs. The approaches to identifying and addressing these needs must also advance, recognising that explanation experiences are subjective, user-centred processes that interact with users towards a better understanding of AI decision-making. This paper delves into the interrelations in multi-faceted XAI and examines how different types of explanations collaboratively meet users' XAI needs. We introduce the Intent Fulfilment Framework (IFF) for creating explanation experiences. The novelty of this paper lies in recognising the importance of "follow-up" on explanations for obtaining clarity, verification and/or substitution. Moreover, the Explanation Experience Dialogue Model integrates the IFF and "Explanation Followups" to provide users with a conversational interface for exploring their explanation needs, thereby creating explanation experiences. Quantitative and qualitative findings from our comparative user study demonstrate the impact of the IFF in improving user engagement, the utility of the AI system and the overall user experience. Overall, we reinforce the principle that "one explanation does not fit all" to create explanation experiences that guide the complex interaction through conversation.

The trustworthiness of AI applications has been the subject of recent research and is also addressed in the EU's recently adopted AI Regulation. The currently emerging foundation models in the field of text, speech and image processing offer completely new possibilities for developing AI applications. This whitepaper shows how the trustworthiness of an AI application developed with foundation models can be evaluated and ensured. For this purpose, the application-specific, risk-based approach for testing and ensuring the trustworthiness of AI applications, as developed in the 'AI Assessment Catalog - Guideline for Trustworthy Artificial Intelligence' by Fraunhofer IAIS, is transferred to the context of foundation models. Special consideration is given to the fact that specific risks of foundation models can have an impact on the AI application and must also be taken into account when checking trustworthiness. Chapter 1 of the white paper explains the fundamental relationship between foundation models and AI applications based on them in terms of trustworthiness. Chapter 2 provides an introduction to the technical construction of foundation models and Chapter 3 shows how AI applications can be developed based on them. Chapter 4 provides an overview of the resulting risks regarding trustworthiness. Chapter 5 shows which requirements for AI applications and foundation models are to be expected according to the draft of the European Union's AI Regulation and Chapter 6 finally shows the system and procedure for meeting trustworthiness requirements.

In this paper, we demonstrate the feasibility of alterfactual explanations for black box image classifiers. Traditional explanation mechanisms from the field of Counterfactual Thinking are a widely-used paradigm for Explainable Artificial Intelligence (XAI), as they follow a natural way of reasoning that humans are familiar with. However, most common approaches from this field are based on communicating information about features or characteristics that are especially important for an AI's decision. However, to fully understand a decision, not only knowledge about relevant features is needed, but the awareness of irrelevant information also highly contributes to the creation of a user's mental model of an AI system. To this end, a novel approach for explaining AI systems called alterfactual explanations was recently proposed on a conceptual level. It is based on showing an alternative reality where irrelevant features of an AI's input are altered. By doing so, the user directly sees which input data characteristics can change arbitrarily without influencing the AI's decision. In this paper, we show for the first time that it is possible to apply this idea to black box models based on neural networks. To this end, we present a GAN-based approach to generate these alterfactual explanations for binary image classifiers. Further, we present a user study that gives interesting insights on how alterfactual explanations can complement counterfactual explanations.

We consider that existing approaches to AI "safety" and "alignment" may not be using the most effective tools, teams, or approaches. We suggest that an alternative and better approach to the problem may be to treat alignment as a social science problem, since the social sciences enjoy a rich toolkit of models for understanding and aligning motivation and behavior, much of which could be repurposed to problems involving AI models, and enumerate reasons why this is so. We introduce an alternate alignment approach informed by social science tools and characterized by three steps: 1. defining a positive desired social outcome for human/AI collaboration as the goal or "North Star," 2. properly framing knowns and unknowns, and 3. forming diverse teams to investigate, observe, and navigate emerging challenges in alignment.

In the digital age, the prevalence of misleading news headlines poses a significant challenge to information integrity, necessitating robust detection mechanisms. This study explores the efficacy of Large Language Models (LLMs) in identifying misleading versus non-misleading news headlines. Utilizing a dataset of 60 articles, sourced from both reputable and questionable outlets across health, science & tech, and business domains, we employ three LLMs- ChatGPT-3.5, ChatGPT-4, and Gemini-for classification. Our analysis reveals significant variance in model performance, with ChatGPT-4 demonstrating superior accuracy, especially in cases with unanimous annotator agreement on misleading headlines. The study emphasizes the importance of human-centered evaluation in developing LLMs that can navigate the complexities of misinformation detection, aligning technical proficiency with nuanced human judgment. Our findings contribute to the discourse on AI ethics, emphasizing the need for models that are not only technically advanced but also ethically aligned and sensitive to the subtleties of human interpretation.

In this study, we aim to identify the most effective machine learning model for accurately classifying Myers-Briggs Type Indicator (MBTI) types from Reddit posts and a Kaggle data set. We apply multi-label classification using the Binary Relevance method. We use Explainable Artificial Intelligence (XAI) approach to highlight the transparency and understandability of the process and result. To achieve this, we experiment with glass-box learning models, i.e. models designed for simplicity, transparency, and interpretability. We selected k-Nearest Neighbour, Multinomial Naive Bayes, and Logistic Regression for the glass-box models. We show that Multinomial Naive Bayes and k-Nearest Neighbour perform better if classes with Observer (S) traits are excluded, whereas Logistic Regression obtains its best results when all classes have > 550 entries.

Does the stethoscope in the picture make the adjacent person a doctor or a patient? This, of course, depends on the contextual relationship of the two objects. If it's obvious, why don't explanation methods for vision models use contextual information? In this paper, we (1) review the most popular methods of explaining computer vision models by pointing out that they do not take into account context information, (2) show examples of failures of popular XAI methods, (3) provide examples of real-world use cases where spatial context plays a significant role, (4) propose new research directions that may lead to better use of context information in explaining computer vision models, (5) argue that a change in approach to explanations is needed from 'where' to 'how'.

Global Concept Explanations for Graphs by Contrastive Learning

Jonas Teufel, Pascal Friederich

Beyond improving trust and validating model fairness, xAI practices also have the potential to recover valuable scientific insights in application domains where little to no prior human intuition exists. To that end, we propose a method to extract global concept explanations from the predictions of graph neural networks to develop a deeper understanding of the tasks underlying structure-property relationships. We identify concept explanations as dense clusters in the self-explaining Megan models subgraph latent space. For each concept, we optimize a representative prototype graph and optionally use GPT-4 to provide hypotheses about why each structure has a certain effect on the prediction. We conduct computational experiments on synthetic and real-world graph property prediction tasks. For the synthetic tasks we find that our method correctly reproduces the structural rules by which they were created. For real-world molecular property regression and classification tasks, we find that our method rediscovers established rules of thumb. More specifically, our results for molecular mutagenicity prediction indicate more fine-grained resolution of structural details than existing explainability methods, consistent with previous results from chemistry literature. Overall, our results show promising capability to extract the underlying structure-property relationships for complex graph property prediction tasks.

Technological advancements have undoubtedly revolutionized various aspects of human life, altering the ways we perceive the world, engage with others, build relationships, and conduct our daily work routines. Among the recent advancements, the proliferation of virtual and mixed reality technologies stands out as a significant leap forward, promising to elevate our experiences and interactions to unprecedented levels. However, alongside the benefits, these emerging technologies also introduce novel avenues for harm and misuse, particularly in virtual and embodied spaces such as Zoom and virtual reality (VR) environments. The immersive nature of virtual reality environments raises unique challenges regarding psychological and emotional well-being. While VR can offer captivating and immersive experiences, prolonged exposure to virtual environments may lead to phenomena like cybersickness, disorientation, and even psychological distress in susceptible individuals. Additionally, the blurring of boundaries between virtual and real-world interactions in VR raises ethical concerns regarding consent, harassment, and the potential for virtual experiences to influence real-life behavior. Additionally, the increasing integration of artificial intelligence (AI) and machine learning algorithms in virtual spaces introduces risks related to algorithmic bias, discrimination, and manipulation. In VR environments, AI-driven systems may inadvertently perpetuate stereotypes, amplify inequalities, or manipulate user behavior through personalized content recommendations and targeted advertising, posing ethical dilemmas and societal risks.

Gaze-based interactions offer a potential way for users to naturally engage with mixed reality (XR) interfaces. Black-box machine learning models enabled higher accuracy for gaze-based interactions. However, due to the black-box nature of the model, users might not be able to understand and effectively adapt their gaze behaviour to achieve high quality interaction. We posit that explainable AI (XAI) techniques can facilitate understanding of and interaction with gaze-based model-driven system in XR. To study this, we built a real-time, multi-level XAI interface for gaze-based interaction using a deep learning model, and evaluated it during a visual search task in XR. A between-subjects study revealed that participants who interacted with XAI made more accurate selections compared to those who did not use the XAI system (i.e., F1 score increase of 10.8%). Additionally, participants who used the XAI system adapted their gaze behavior over time to make more effective selections. These findings suggest that XAI can potentially be used to assist users in more effective collaboration with model-driven interactions in XR.

Deep learning is dramatically transforming the field of medical imaging and radiology, enabling the identification of pathologies in medical images, including computed tomography (CT) and X-ray scans. However, the performance of deep learning models, particularly in segmentation tasks, is often limited by the need for extensive annotated datasets. To address this challenge, the capabilities of weakly supervised semantic segmentation are explored through the lens of Explainable AI and the generation of counterfactual explanations. The scope of this research is development of a novel counterfactual inpainting approach (COIN) that flips the predicted classification label from abnormal to normal by using a generative model. For instance, if the classifier deems an input medical image X as abnormal, indicating the presence of a pathology, the generative model aims to inpaint the abnormal region, thus reversing the classifier's original prediction label. The approach enables us to produce precise segmentations for pathologies without depending on pre-existing segmentation masks. Crucially, image-level labels are utilized, which are substantially easier to acquire than creating detailed segmentation masks. The effectiveness of the method is demonstrated by segmenting synthetic targets and actual kidney tumors from CT images acquired from Tartu University Hospital in Estonia. The findings indicate that COIN greatly surpasses established attribution methods, such as RISE, ScoreCAM, and LayerCAM, as well as an alternative counterfactual explanation method introduced by Singla et al. This evidence suggests that COIN is a promising approach for semantic segmentation of tumors in CT images, and presents a step forward in making deep learning applications more accessible and effective in healthcare, where annotated data is scarce.

The worldwide adoption of machine learning (ML) and deep learning models, particularly in critical sectors, such as healthcare and finance, presents substantial challenges in maintaining individual privacy and fairness. These two elements are vital to a trustworthy environment for learning systems. While numerous studies have concentrated on protecting individual privacy through differential privacy (DP) mechanisms, emerging research indicates that differential privacy in machine learning models can unequally impact separate demographic subgroups regarding prediction accuracy. This leads to a fairness concern, and manifests as biased performance. Although the prevailing view is that enhancing privacy intensifies fairness disparities, a smaller, yet significant, subset of research suggests the opposite view. In this article, with extensive evaluation results, we demonstrate that the impact of differential privacy on fairness is not monotonous. Instead, we observe that the accuracy disparity initially grows as more DP noise (enhanced privacy) is added to the ML process, but subsequently diminishes at higher privacy levels with even more noise. Moreover, implementing gradient clipping in the differentially private stochastic gradient descent ML method can mitigate the negative impact of DP noise on fairness. This mitigation is achieved by moderating the disparity growth through a lower clipping threshold.