Previous research in Explainable Artificial Intelligence (XAI) suggests that a main aim of explainability approaches is to satisfy specific interests, goals, expectations, needs, and demands regarding artificial systems (we call these stakeholders' desiderata) in a variety of contexts. However, the literature on XAI is vast, spreads out across multiple largely disconnected disciplines, and it often remains unclear how explainability approaches are supposed to achieve the goal of satisfying stakeholders' desiderata. This paper discusses the main classes of stakeholders calling for explainability of artificial systems and reviews their desiderata. We provide a model that explicitly spells out the main concepts and relations necessary to consider and investigate when evaluating, adjusting, choosing, and developing explainability approaches that aim to satisfy stakeholders' desiderata. This model can serve researchers from the variety of different disciplines involved in XAI as a common ground. It emphasizes where there is interdisciplinary potential in the evaluation and the development of explainability approaches.

This graduate textbook on machine learning tells a story of how patterns in data support predictions and consequential actions. Starting with the foundations of decision making, we cover representation, optimization, and generalization as the constituents of supervised learning. A chapter on datasets as benchmarks examines their histories and scientific bases. Self-contained introductions to causality, the practice of causal inference, sequential decision making, and reinforcement learning equip the reader with concepts and tools to reason about actions and their consequences. Throughout, the text discusses historical context and societal impact. We invite readers from all backgrounds; some experience with probability, calculus, and linear algebra suffices.

Cyber-defense systems are being developed to automatically ingest Cyber Threat Intelligence (CTI) that contains semi-structured data and/or text to populate knowledge graphs. A potential risk is that fake CTI can be generated and spread through Open-Source Intelligence (OSINT) communities or on the Web to effect a data poisoning attack on these systems. Adversaries can use fake CTI examples as training input to subvert cyber defense systems, forcing the model to learn incorrect inputs to serve their malicious needs. In this paper, we automatically generate fake CTI text descriptions using transformers. We show that given an initial prompt sentence, a public language model like GPT-2 with fine-tuning, can generate plausible CTI text with the ability of corrupting cyber-defense systems. We utilize the generated fake CTI text to perform a data poisoning attack on a Cybersecurity Knowledge Graph (CKG) and a cybersecurity corpus. The poisoning attack introduced adverse impacts such as returning incorrect reasoning outputs, representation poisoning, and corruption of other dependent AI-based cyber defense systems. We evaluate with traditional approaches and conduct a human evaluation study with cybersecurity professionals and threat hunters. Based on the study, professional threat hunters were equally likely to consider our fake generated CTI as true.

In this paper I explore the scaffolding of normative assumptions that supports Sabina Leonelli's implicit appeal to the values of epistemic integrity and the global public good that conjointly animate the ethos of responsible and sustainable data work in the context of COVID-19. Drawing primarily on the writings of sociologist Robert K. Merton, the thinkers of the Vienna Circle, and Charles Sanders Peirce, I make some of these assumptions explicit by telling a longer story about the evolution of social thinking about the normative structure of science from Merton's articulation of his well-known norms (those of universalism, communism, organized skepticism, and disinterestedness) to the present. I show that while Merton's norms and his intertwinement of these with the underlying mechanisms of democratic order provide us with an especially good starting point to explore and clarify the commitments and values of science, Leonelli's broader, more context-responsive, and more holistic vision of the epistemic integrity of data scientific understanding, and her discernment of the global and biospheric scope of its moral-practical reach, move beyond Merton's schema in ways that effectively draw upon important critiques. Stepping past Merton, I argue that a combination of situated universalism, methodological pluralism, strong objectivity, and unbounded communalism must guide the responsible and sustainable data work of the future.

Our work represents another step into the detection and prevention of these ever-more present political manipulation efforts. We, therefore, start by focusing on understanding what the state-of-the-art approaches lack -- since the problem remains, this is a fair assumption. We find concerning issues within the current literature and follow a diverging path. Notably, by placing emphasis on using data features that are less susceptible to malicious manipulation and also on looking for high-level approaches that avoid a granularity level that is biased towards easy-to-spot and low impact cases. We designed and implemented a framework -- Twitter Watch -- that performs structured Twitter data collection, applying it to the Portuguese Twittersphere. We investigate a data snapshot taken on May 2020, with around 5 million accounts and over 120 million tweets (this value has since increased to over 175 million). The analyzed time period stretches from August 2019 to May 2020, with a focus on the Portuguese elections of October 6th, 2019. However, the Covid-19 pandemic showed itself in our data, and we also delve into how it affected typical Twitter behavior. We performed three main approaches: content-oriented, metadata-oriented, and network interaction-oriented. We learn that Twitter's suspension patterns are not adequate to the type of political trolling found in the Portuguese Twittersphere -- identified by this work and by an independent peer - nor to fake news posting accounts. We also surmised that the different types of malicious accounts we independently gathered are very similar both in terms of content and interaction, through two distinct analysis, and are simultaneously very distinct from regular accounts.

The optimization of reactions used to synthesize target compounds is pivotal to chemical research and discovery, whether in developing a route for manufacturing a life-saving medicine1 or unlocking the potential of a new material2. But reaction optimization requires iterative experiments to balance the often conflicting effects of numerous coupled variables, and frequently involves finding the sweet spot among thousands of possible sets of experimental conditions. Expert synthetic chemists currently navigate this expansive experimental void using simplified model reactions, heuristic approaches and intuition derived from observation of experimental data3. Writing in Nature, Shields et al.4 report machine-learning software that can optimize diverse classes of reaction with fewer iterations, on average, than are needed by humans. Machine learning has emerged as a useful tool for various aspects of chemical synthesis, because it is ideally suited to extrapolating predictive models that are used to solve synthetic problems by recognizing patterns in multidimensional data sets5.

While algorithm audits are growing rapidly in commonality and public importance, relatively little scholarly work has gone toward synthesizing prior work and strategizing future research in the area. This systematic literature review aims to do just that, following PRISMA guidelines in a review of over 500 English articles that yielded 62 algorithm audit studies. The studies are synthesized and organized primarily by behavior (discrimination, distortion, exploitation, and misjudgement), with codes also provided for domain (e.g. search, vision, advertising, etc.), organization (e.g. Google, Facebook, Amazon, etc.), and audit method (e.g. sock puppet, direct scrape, crowdsourcing, etc.). The review shows how previous audit studies have exposed public-facing algorithms exhibiting problematic behavior, such as search algorithms culpable of distortion and advertising algorithms culpable of discrimination. Based on the studies reviewed, it also suggests some behaviors (e.g. discrimination on the basis of intersectional identities), domains (e.g. advertising algorithms), methods (e.g. code auditing), and organizations (e.g. Twitter, TikTok, LinkedIn) that call for future audit attention. The paper concludes by offering the common ingredients of successful audits, and discussing algorithm auditing in the context of broader research working toward algorithmic justice.

New drone video shows the recent damage wrought on California's iconic Highway 1, where part of the road collapsed after heavy rains washed it into the ocean last week. The video, released by the Monterey County Sheriff's Office, shows a large part of the highway still flooded and covered with debris from recent rainfall and mudslides. At the point of collapse, about 45 miles south of Carmel in the Big Sur area, both lanes of the road are completely gone, with a massive hole sloping toward the Pacific Ocean in its place. The sheriff's office video shows water running through the collapsed part of the road, which by Friday had fallen into the sea. California has been plagued by extensive mudslides, largely in areas burned out during the previous season's wildfires.

The growing energy and performance costs of deep learning have driven the community to reduce the size of neural networks by selectively pruning components. Similarly to their biological counterparts, sparse networks generalize just as well, if not better than, the original dense networks. Sparsity can reduce the memory footprint of regular networks to fit mobile devices, as well as shorten training time for ever growing networks. In this paper, we survey prior work on sparsity in deep learning and provide an extensive tutorial of sparsification for both inference and training. We describe approaches to remove and add elements of neural networks, different training strategies to achieve model sparsity, and mechanisms to exploit sparsity in practice. Our work distills ideas from more than 300 research papers and provides guidance to practitioners who wish to utilize sparsity today, as well as to researchers whose goal is to push the frontier forward. We include the necessary background on mathematical methods in sparsification, describe phenomena such as early structure adaptation, the intricate relations between sparsity and the training process, and show techniques for achieving acceleration on real hardware. We also define a metric of pruned parameter efficiency that could serve as a baseline for comparison of different sparse networks. We close by speculating on how sparsity can improve future workloads and outline major open problems in the field.

After challenging the validity of these assumptions in real-world applications, we propose ways to move forward when they are violated. First, we show that group fairness criteria purely based on statistical properties of observed data are fundamentally limited. Revisiting this limitation from a causal viewpoint we develop a more versatile conceptual framework, causal fairness criteria, and first algorithms to achieve them. We also provide tools to analyze how sensitive a believed-to-be causally fair algorithm is to misspecifications of the causal graph. Second, we overcome the assumption that sensitive data is readily available in practice. To this end we devise protocols based on secure multi-party computation to train, validate, and contest fair decision algorithms without requiring users to disclose their sensitive data or decision makers to disclose their models. Finally, we also accommodate the fact that outcome labels are often only observed when a certain decision has been made. We suggest a paradigm shift away from training predictive models towards directly learning decisions to relax the traditional assumption that labels can always be recorded. The main contribution of this thesis is the development of theoretically substantiated and practically feasible methods to move research on fair machine learning closer to real-world applications.