Research Papers

Models trained with empirical risk minimization (ERM) are known to learn to rely on spurious features, i.e., their prediction is based on undesired auxiliary features which are strongly correlated with class labels but lack causal reasoning. This behavior particularly degrades accuracy in groups of samples of the correlated class that are missing the spurious feature or samples of the opposite class but with the spurious feature present. The recently proposed Deep Feature Reweighting (DFR) method improves accuracy of these worst groups. Based on the main argument that ERM mods can learn core features sufficiently well, DFR only needs to retrain the last layer of the classification model with a small group-balanced data set. In this work, we examine the applicability of DFR to realistic data in the medical domain. Furthermore, we investigate the reasoning behind the effectiveness of last-layer retraining and show that even though DFR has the potential to improve the accuracy of the worst group, it remains susceptible to spurious correlations.

Designing a Direct Feedback Loop between Humans and Convolutional Neural Networks through Local Explanations

Tong Steven Sun, Yuyang Gao, Shubham Khaladkar +4

The local explanation provides heatmaps on images to explain how Convolutional Neural Networks (CNNs) derive their output. Due to its visual straightforwardness, the method has been one of the most popular explainable AI (XAI) methods for diagnosing CNNs. Through our formative study (S1), however, we captured ML engineers' ambivalent perspective about the local explanation as a valuable and indispensable envision in building CNNs versus the process that exhausts them due to the heuristic nature of detecting vulnerability. Moreover, steering the CNNs based on the vulnerability learned from the diagnosis seemed highly challenging. To mitigate the gap, we designed DeepFuse, the first interactive design that realizes the direct feedback loop between a user and CNNs in diagnosing and revising CNN's vulnerability using local explanations. DeepFuse helps CNN engineers to systemically search "unreasonable" local explanations and annotate the new boundaries for those identified as unreasonable in a labor-efficient manner. Next, it steers the model based on the given annotation such that the model doesn't introduce similar mistakes. We conducted a two-day study (S2) with 12 experienced CNN engineers. Using DeepFuse, participants made a more accurate and "reasonable" model than the current state-of-the-art. Also, participants found the way DeepFuse guides case-based reasoning can practically improve their current practice. We provide implications for design that explain how future HCI-driven design can move our practice forward to make XAI-driven insights more actionable.

Agency and legibility for artists through Experiential AI

Drew Hemment, Matjaz Vidmar, Daga Panas +3

Experiential AI is an emerging research field that addresses the challenge of making AI tangible and explicit, both to fuel cultural experiences for audiences, and to make AI systems more accessible to human understanding. The central theme is how artists, scientists and other interdisciplinary actors can come together to understand and communicate the functionality of AI, ML and intelligent robots, their limitations, and consequences, through informative and compelling experiences. It provides an approach and methodology for the arts and tangible experiences to mediate between impenetrable computer code and human understanding, making not just AI systems but also their values and implications more transparent, and therefore accountable. In this paper, we report on an empirical case study of an experiential AI system designed for creative data exploration of a user-defined dimension, to enable creators to gain more creative control over the AI process. We discuss how experiential AI can increase legibility and agency for artists, and how the arts can provide creative strategies and methods which can add to the toolbox for human-centred XAI.

The rapid advancement of Artificial Intelligence (AI), represented by ChatGPT, has raised concerns about responsible AI development and utilization. Existing frameworks lack a comprehensive synthesis of AI risk assessment questions. To address this, we introduce QB4AIRA, a novel question bank developed by refining questions from five globally recognized AI risk frameworks, categorized according to Australia's AI ethics principles. QB4AIRA comprises 293 prioritized questions covering a wide range of AI risk areas, facilitating effective risk assessment. It serves as a valuable resource for stakeholders in assessing and managing AI risks, while paving the way for new risk frameworks and guidelines. By promoting responsible AI practices, QB4AIRA contributes to responsible AI deployment, mitigating potential risks and harms.

This paper reports on our initial evaluation of The Equitable AI Research Roundtable -- a coalition of experts in law, education, community engagement, social justice, and technology. EARR was created in collaboration among a large tech firm, nonprofits, NGO research institutions, and universities to provide critical research based perspectives and feedback on technology's emergent ethical and social harms. Through semi-structured workshops and discussions within the large tech firm, EARR has provided critical perspectives and feedback on how to conceptualize equity and vulnerability as they relate to AI technology. We outline three principles in practice of how EARR has operated thus far that are especially relevant to the concerns of the FAccT community: how EARR expands the scope of expertise in AI development, how it fosters opportunities for epistemic curiosity and responsibility, and that it creates a space for mutual learning. This paper serves as both an analysis and translation of lessons learned through this engagement approach, and the possibilities for future research.

Recurrent Neural Networks (RNNs) have been widely used in Natural Language Processing (NLP) tasks given its superior performance on processing sequential data. However, it is challenging to interpret and debug RNNs due to the inherent complexity and the lack of transparency of RNNs. While many explainable AI (XAI) techniques have been proposed for RNNs, most of them only support local explanations rather than global explanations. In this paper, we present DeepSeer, an interactive system that provides both global and local explanations of RNN behavior in multiple tightly-coordinated views for model understanding and debugging. The core of DeepSeer is a state abstraction method that bundles semantically similar hidden states in an RNN model and abstracts the model as a finite state machine. Users can explore the global model behavior by inspecting text patterns associated with each state and the transitions between states. Users can also dive into individual predictions by inspecting the state trace and intermediate prediction results of a given input. A between-subjects user study with 28 participants shows that, compared with a popular XAI technique, LIME, participants using DeepSeer made a deeper and more comprehensive assessment of RNN model behavior, identified the root causes of incorrect predictions more accurately, and came up with more actionable plans to improve the model performance.

Maximizing a submodular function has a wide range of applications in machine learning and data mining. One such application is data summarization whose goal is to select a small set of representative and diverse data items from a large dataset. However, data items might have sensitive attributes such as race or gender, in this setting, it is important to design \emph{fairness-aware} algorithms to mitigate potential algorithmic bias that may cause over- or under- representation of particular groups. Motivated by that, we propose and study the classic non-monotone submodular maximization problem subject to novel group fairness constraints. Our goal is to select a set of items that maximizes a non-monotone submodular function, while ensuring that the number of selected items from each group is proportionate to its size, to the extent specified by the decision maker. We develop the first constant-factor approximation algorithms for this problem. We also extend the basic model to incorporate an additional global size constraint on the total number of selected items.

We consider the problem EnumIP of enumerating prime implicants of Boolean functions represented by decision decomposable negation normal form (dec-DNNF) circuits. We study EnumIP from dec-DNNF within the framework of enumeration complexity and prove that it is in OutputP, the class of output polynomial enumeration problems, and more precisely in IncP, the class of polynomial incremental time enumeration problems. We then focus on two closely related, but seemingly harder, enumeration problems where further restrictions are put on the prime implicants to be generated. In the first problem, one is only interested in prime implicants representing subset-minimal abductive explanations, a notion much investigated in AI for more than three decades. In the second problem, the target is prime implicants representing sufficient reasons, a recent yet important notion in the emerging field of eXplainable AI, since they aim to explain predictions achieved by machine learning classifiers. We provide evidence showing that enumerating specific prime implicants corresponding to subset-minimal abductive explanations or to sufficient reasons is not in OutputP.

Handwritten character recognition is a hot topic for research nowadays. If we can convert a handwritten piece of paper into a text-searchable document using the Optical Character Recognition (OCR) technique, we can easily understand the content and do not need to read the handwritten document. OCR in the English language is very common, but in the Bengali language, it is very hard to find a good quality OCR application. If we can merge machine learning and deep learning with OCR, it could be a huge contribution to this field. Various researchers have proposed a number of strategies for recognizing Bengali handwritten characters. A lot of ML algorithms and deep neural networks were used in their work, but the explanations of their models are not available. In our work, we have used various machine learning algorithms and CNN to recognize handwritten Bengali digits. We have got acceptable accuracy from some ML models, and CNN has given us great testing accuracy. Grad-CAM was used as an XAI method on our CNN model, which gave us insights into the model and helped us detect the origin of interest for recognizing a digit from an image.

The applications of Artificial Intelligence (AI) surround decisions on increasingly many aspects of human lives. Society responds by imposing legal and social expectations for the accountability of such automated decision systems (ADSs). Fairness, a fundamental constituent of AI accountability, is concerned with just treatment of individuals and sensitive groups (e.g., based on sex, race). While many studies focus on fair learning and fairness testing for the classification tasks, the literature is rather limited on how to examine fairness in regression tasks. This work presents error parity as a regression fairness notion and introduces a testing methodology to assess group fairness based on a statistical hypothesis testing procedure. The error parity test checks whether prediction errors are distributed similarly across sensitive groups to determine if an ADS is fair. It is followed by a suitable permutation test to compare groups on several statistics to explore disparities and identify impacted groups. The usefulness and applicability of the proposed methodology are demonstrated via a case study on COVID-19 projections in the US at the county level, which revealed race-based differences in forecast errors. Overall, the proposed regression fairness testing methodology fills a gap in the fair machine learning literature and may serve as a part of larger accountability assessments and algorithm audits.

The new era of image segmentation leveraging the power of Deep Neural Nets (DNNs) comes with a price tag: to train a neural network for pixel-wise segmentation, a large amount of training samples has to be manually labeled on pixel-precision. In this work, we address this by following an indirect solution. We build upon the advances of the Explainable AI (XAI) community and extract a pixel-wise binary segmentation from the output of the Layer-wise Relevance Propagation (LRP) explaining the decision of a classification network. We show that we achieve similar results compared to an established U-Net segmentation architecture, while the generation of the training data is significantly simplified. The proposed method can be trained in a weakly supervised fashion, as the training samples must be only labeled on image-level, at the same time enabling the output of a segmentation mask. This makes it especially applicable to a wider range of real applications where tedious pixel-level labelling is often not possible.

Monumental advances in deep learning have led to unprecedented achievements across various domains. While the performance of deep neural networks is indubitable, the architectural design and interpretability of such models are nontrivial. Research has been introduced to automate the design of neural network architectures through neural architecture search (NAS). Recent progress has made these methods more pragmatic by exploiting distributed computation and novel optimization algorithms. However, there is little work in optimizing architectures for interpretability. To this end, we propose a multi-objective distributed NAS framework that optimizes for both task performance and "introspectability," a surrogate metric for aspects of interpretability. We leverage the non-dominated sorting genetic algorithm (NSGA-II) and explainable AI (XAI) techniques to reward architectures that can be better comprehended by domain experts. The framework is evaluated on several image classification datasets. We demonstrate that jointly optimizing for task error and introspectability leads to more disentangled and debuggable architectures that perform within tolerable error.