We consider the challenging problem of using domain knowledge to improve deep reinforcement learning policies. To this end, we propose LEGIBLE, a novel approach, following a multi-step process, which starts by mining rules from a deep RL policy, constituting a partially symbolic representation. These rules describe which decisions the RL policy makes and which it avoids making. In the second step, we generalize the mined rules using domain knowledge expressed as metamorphic relations. We adapt these relations from software testing to RL to specify expected changes of actions in response to changes in observations. The third step is evaluating generalized rules to determine which generalizations improve performance when enforced. These improvements show weaknesses in the policy, where it has not learned the general rules and thus can be improved by rule guidance. LEGIBLE supported by metamorphic relations provides a principled way of expressing and enforcing domain knowledge about RL environments. We show the efficacy of our approach by demonstrating that it effectively finds weaknesses, accompanied by explanations of these weaknesses, in eleven RL environments and by showcasing that guiding policy execution with rules improves performance w.r.t. gained reward.

Akin to neuroplasticity in human brains, the plasticity of deep neural networks enables their quick adaption to new data. This makes plasticity particularly crucial for deep Reinforcement Learning (RL) agents: Once plasticity is lost, an agent's performance will inevitably plateau because it cannot improve its policy to account for changes in the data distribution, which are a necessary consequence of its learning process. Thus, developing well-performing and sample-efficient agents hinges on their ability to remain plastic during training. Furthermore, the loss of plasticity can be connected to many other issues plaguing deep RL, such as training instabilities, scaling failures, overestimation bias, and insufficient exploration. With this survey, we aim to provide an overview of the emerging research on plasticity loss for academics and practitioners of deep reinforcement learning. First, we propose a unified definition of plasticity loss based on recent works, relate it to definitions from the literature, and discuss metrics for measuring plasticity loss. Then, we categorize and discuss numerous possible causes of plasticity loss before reviewing currently employed mitigation strategies. Our taxonomy is the first systematic overview of the current state of the field. Lastly, we discuss prevalent issues within the literature, such as a necessity for broader evaluation, and provide recommendations for future research, like gaining a better understanding of an agent's neural activity and behavior.

This work investigates an important phenomenon in centroid-based deep clustering (DC) algorithms: Performance quickly saturates after a period of rapid early gains. Practitioners commonly address early saturation with periodic reclustering, which we demonstrate to be insufficient to address performance plateaus. We call this phenomenon the "reclustering barrier" and empirically show when the reclustering barrier occurs, what its underlying mechanisms are, and how it is possible to Break the Reclustering Barrier with our algorithm BRB. BRB avoids early over-commitment to initial clusterings and enables continuous adaptation to reinitialized clustering targets while remaining conceptually simple. Applying our algorithm to widely-used centroid-based DC algorithms, we show that (1) BRB consistently improves performance across a wide range of clustering benchmarks, (2) BRB enables training from scratch, and (3) BRB performs competitively against state-of-the-art DC algorithms when combined with a contrastive loss. We release our code and pre-trained models at https://github.com/Probabilistic-and-Interactive-ML/breaking-the-reclustering-barrier .

We discuss the role of humans in algorithmic decision-making (ADM) for socially relevant problems from a technical and philosophical perspective. In particular, we illustrate tensions arising from diverse expectations, values, and constraints by and on the humans involved. To this end, we assume that a strategic decision-maker (SDM) introduces ADM to optimize strategic and societal goals while the algorithms' recommended actions are overseen by a practical decision-maker (PDM) - a specific human-in-the-loop - who makes the final decisions. While the PDM is typically assumed to be a corrective, it can counteract the realization of the SDM's desired goals and societal values not least because of a misalignment of these values and unmet information needs of the PDM. This has significant implications for the distribution of power between the stakeholders in ADM, their constraints, and information needs. In particular, we emphasize the overseeing PDM's role as a potential political and ethical decision maker, who acts expected to balance strategic, value-driven objectives and on-the-ground individual decisions and constraints. We demonstrate empirically, on a machine learning benchmark dataset, the significant impact an overseeing PDM's decisions can have even if the PDM is constrained to performing only a limited amount of actions differing from the algorithms' recommendations. To ensure that the SDM's intended values are realized, the PDM needs to be provided with appropriate information conveyed through tailored explanations and its role must be characterized clearly. Our findings emphasize the need for an in-depth discussion of the role and power of the PDM and challenge the often-taken view that just including a human-in-the-loop in ADM ensures the 'correct' and 'ethical' functioning of the system.

Explanations of AI systems rarely address the information needs of people affected by algorithmic decision-making (ADM). This gap between conveyed information and information that matters to affected stakeholders can impede understanding and adherence to regulatory frameworks such as the AI Act. To address this gap, we present the "XAI Novice Question Bank": A catalog of affected stakeholders' information needs in two ADM use cases (employment prediction and health monitoring), covering the categories data, system context, system usage, and system specifications. Information needs were gathered in an interview study where participants received explanations in response to their inquiries. Participants further reported their understanding and decision confidence, showing that while confidence tended to increase after receiving explanations, participants also met understanding challenges, such as being unable to tell why their understanding felt incomplete. Explanations further influenced participants' perceptions of the systems' risks and benefits, which they confirmed or changed depending on the use case. When risks were perceived as high, participants expressed particular interest in explanations about intention, such as why and to what end a system was put in place. With this work, we aim to support the inclusion of affected stakeholders into explainability by contributing an overview of information and challenges relevant to them when deciding on the adoption of ADM systems. We close by summarizing our findings in a list of six key implications that inform the design of future explanations for affected stakeholder audiences.

We consider the problem of learning Variational Autoencoders (VAEs), i.e., a type of deep generative model, from data with missing values. Such data is omnipresent in real-world applications of machine learning because complete data is often impossible or too costly to obtain. We particularly focus on improving a VAE's amortized posterior inference, i.e., the encoder, which in the case of missing data can be susceptible to learning inconsistent posterior distributions regarding the missingness. To this end, we provide a formal definition of posterior consistency and propose an approach for regularizing an encoder's posterior distribution which promotes this consistency. We observe that the proposed regularization suggests a different training objective than that typically considered in the literature when facing missing values. Furthermore, we empirically demonstrate that our regularization leads to improved performance in missing value settings in terms of reconstruction quality and downstream tasks utilizing uncertainty in the latent space. This improved performance can be observed for many classes of VAEs including VAEs equipped with normalizing flows.

Student modeling is central to many educational technologies as it enables the prediction of future learning outcomes and targeted instructional strategies. However, open-ended learning environments pose challenges for accurately modeling students due to the diverse behaviors exhibited by students and the absence of a well-defined set of learning skills. To approach these challenges, we explore the application of Large Language Models (LLMs) for in-context student modeling in open-ended learning environments. We introduce a novel framework, LLM-SS, that leverages LLMs for synthesizing student's behavior. More concretely, given a particular student's solving attempt on a reference task as observation, the goal is to synthesize the student's attempt on a target task. Our framework can be combined with different LLMs; moreover, we fine-tune LLMs using domain-specific expertise to boost their understanding of domain background and student behaviors. We evaluate several concrete methods based on LLM-SS using the StudentSyn benchmark, an existing student's attempt synthesis benchmark in visual programming. Experimental results show a significant improvement compared to baseline methods included in the StudentSyn benchmark. Furthermore, our method using the fine-tuned Llama2-70B model improves noticeably compared to using the base model and becomes on par with using the state-of-the-art GPT-4 model.

Well-known examples of such "high-risk" [6] systems can be found in recidivism prediction [5], refugee resettlement [3], and public employment [19]. Many authors have outlined that faulty or biased predictions by ADM systems can have far-reaching consequences, including discrimination [5], inaccurate predictions [4], and overreliance on automated decisions [2]. Therefore, high-level guidelines are meant to prevent these issues by pointing out ways to develop trustworthy and ethical AI [10, 22]. However, practically applying these guidelines remains challenging, since the meaning and priority of ethical values shift depending on who is asked [11]. Recent work in Explainable Artificial Intelligence (XAI) thus suggests equipping individuals who are involved with an ADM system and carry responsibility--so-called "stakeholders"--with the means of assessing the system themselves, i.e. enabling users, deployers, and affected individuals to independently check the system's ethical values [14]. Arguably, a pronounced understanding of the system is necessary for making such an assessment. While numerous XAI studies have examined how explaining an ADM system can increase stakeholders' understanding [20, 21], we highlight two aspects that remain an open challenge: i) the amounts of resources needed to produce and test domain-specific explanations and ii) the difficulty of creating and evaluating understanding for a large variety of people. Further, it is important to note that, despite our reference to "Explainable AI," ADM is not constrained to AI, and indeed might encompass a broader problem space. Despite the emphasis on "understanding" in XAI research, the field features only a few studies that introduce learning frameworks from other disciplines.

We propose Convex Constraint Learning for Reinforcement Learning (CoCoRL), a novel approach for inferring shared constraints in a Constrained Markov Decision Process (CMDP) from a set of safe demonstrations with possibly different reward functions. While previous work is limited to demonstrations with known rewards or fully known environment dynamics, CoCoRL can learn constraints from demonstrations with different unknown rewards without knowledge of the environment dynamics. CoCoRL constructs a convex safe set based on demonstrations, which provably guarantees safety even for potentially sub-optimal (but safe) demonstrations. For near-optimal demonstrations, CoCoRL converges to the true safe set with no policy regret. We evaluate CoCoRL in tabular environments and a continuous driving simulation with multiple constraints. CoCoRL learns constraints that lead to safe driving behavior and that can be transferred to different tasks and environments. In contrast, alternative methods based on Inverse Reinforcement Learning (IRL) often exhibit poor performance and learn unsafe policies.

Block-based programming environments are increasingly used to introduce computing concepts to beginners. However, novice students often struggle in these environments, given the conceptual and open-ended nature of programming tasks. To effectively support a student struggling to solve a given task, it is important to provide adaptive scaffolding that guides the student towards a solution. We introduce a scaffolding framework based on pop quizzes presented as multi-choice programming tasks. To automatically generate these pop quizzes, we propose a novel algorithm, PQuizSyn. More formally, given a reference task with a solution code and the student's current attempt, PQuizSyn synthesizes new tasks for pop quizzes with the following features: (a) Adaptive (i.e., individualized to the student's current attempt), (b) Comprehensible (i.e., easy to comprehend and solve), and (c) Concealing (i.e., do not reveal the solution code). Our algorithm synthesizes these tasks using techniques based on symbolic reasoning and graph-based code representations. We show that our algorithm can generate hundreds of pop quizzes for different student attempts on reference tasks from Hour of Code: Maze Challenge and Karel. We assess the quality of these pop quizzes through expert ratings using an evaluation rubric. Further, we have built an online platform for practicing block-based programming tasks empowered via pop quiz based feedback, and report results from an initial user study.