The strong performance (Chowdhery et al., 2022; Nostalgebraist, 2022; OpenAI, 2023; Google, 2023), and emergent abilities (Wei et al., 2022) of modern language models (LMs) depend on self-supervised pretraining on massive text datasets. All model developers implicitly or explicitly decide the composition of these datasets: what data sources to include, whether to filter for attributes such as quality and toxicity, and when to gather new documents. While many of the most prominent models do not document their curation procedures (OpenAI, 2023; Google, 2023), or only document which procedures they used (Brown et al., 2020; Nostalgebraist, 2022; Scao et al., 2022; Touvron et al., 2023), they rarely document why they chose those protocols or what effect they had. This documentation debt leaves practitioners to be guided by intuitions and precedents, neither thoroughly evaluated (Bandy and Vincent, 2021; Sambasivan et al., 2021). Given the outsized and fundamental role of pretraining data in modern LMs, we believe this neglectful practice has detracted from responsible data use and hampered effective model development (Rogers, 2021; Gebru et al., 2021; Bender and Friedman, 2018). Among the small number of general-purpose LMs dominating community use and discussion, the prevailing focus has been on the scale of pretraining data and number of optimization steps (Brown et al., 2020; Nostalgebraist, 2022; Google, 2023). In this work, we systematically test how common data design decisions affect model performance--specifically: the time of collection, content filtering strategy (toxicity/quality), and domain composition. We study the impacts in two ways.

Providing system-generated explanations for recommendations represents an important step towards transparent and trustworthy recommender systems. Explainable recommender systems provide a human-understandable rationale for their outputs. Over the last two decades, explainable recommendation has attracted much attention in the recommender systems research community. This paper aims to provide a comprehensive review of research efforts on visual explanation in recommender systems. More concretely, we systematically review the literature on explanations in recommender systems based on four dimensions, namely explanation goal, explanation scope, explanation style, and explanation format. Recognizing the importance of visualization, we approach the recommender system literature from the angle of explanatory visualizations, that is using visualizations as a display style of explanation. As a result, we derive a set of guidelines that might be constructive for designing explanatory visualizations in recommender systems and identify perspectives for future work in this field. The aim of this review is to help recommendation researchers and practitioners better understand the potential of visually explainable recommendation research and to support them in the systematic design of visual explanations in current and future recommender systems.

This paper presents a policy parameterization for learning-based control on nonlinear, partially-observed dynamical systems. The parameterization is based on a nonlinear version of the Youla parameterization and the recently proposed Recurrent Equilibrium Network (REN) class of models. We prove that the resulting Youla-REN parameterization automatically satisfies stability (contraction) and user-tunable robustness (Lipschitz) conditions on the closed-loop system. This means it can be used for safe learning-based control with no additional constraints or projections required to enforce stability or robustness. We test the new policy class in simulation on two reinforcement learning tasks: 1) magnetic suspension, and 2) inverting a rotary-arm pendulum. We find that the Youla-REN performs similarly to existing learning-based and optimal control methods while also ensuring stability and exhibiting improved robustness to adversarial disturbances.

Knowledge graph embedding (KGE) that maps entities and relations into vector representations is essential for downstream applications. Conventional KGE methods require high-dimensional representations to learn the complex structure of knowledge graph, but lead to oversized model parameters. Recent advances reduce parameters by low-dimensional entity representations, while developing techniques (e.g., knowledge distillation or reinvented representation forms) to compensate for reduced dimension. However, such operations introduce complicated computations and model designs that may not benefit large knowledge graphs. To seek a simple strategy to improve the parameter efficiency of conventional KGE models, we take inspiration from that deeper neural networks require exponentially fewer parameters to achieve expressiveness comparable to wider networks for compositional structures. We view all entity representations as a single-layer embedding network, and conventional KGE methods that adopt high-dimensional entity representations equal widening the embedding network to gain expressiveness. To achieve parameter efficiency, we instead propose a deeper embedding network for entity representations, i.e., a narrow entity embedding layer plus a multi-layer dimension lifting network (LiftNet). Experiments on three public datasets show that by integrating LiftNet, four conventional KGE methods with 16-dimensional representations achieve comparable link prediction accuracy as original models that adopt 512-dimensional representations, saving 68.4% to 96.9% parameters.

When used with deep learning, the symbolic music modality is often coupled with language model architectures. To do so, the music needs to be tokenized, i.e. converted into a sequence of discrete tokens. This can be achieved by different approaches, as music can be composed of simultaneous tracks, of simultaneous notes with several attributes. Until now, the proposed tokenizations rely on small vocabularies of tokens describing the note attributes and time events, resulting in fairly long token sequences, and a sub-optimal use of the embedding space of language models. Recent research has put efforts on reducing the overall sequence length by merging embeddings or combining tokens. In this paper, we show that Byte Pair Encoding, a compression technique widely used for natural language, significantly decreases the sequence length while increasing the vocabulary size. By doing so, we leverage the embedding capabilities of such models with more expressive tokens, resulting in both better results and faster inference in generation and classification tasks. The source code is shared on Github, along with a companion website. Finally, BPE is directly implemented in MidiTok, allowing the reader to easily benefit from this method.

We study fair multi-objective reinforcement learning in which an agent must learn a policy that simultaneously achieves high reward on multiple dimensions of a vector-valued reward. Motivated by the fair resource allocation literature, we model this as an expected welfare maximization problem, for some nonlinear fair welfare function of the vector of long-term cumulative rewards. One canonical example of such a function is the Nash Social Welfare, or geometric mean, the log transform of which is also known as the Proportional Fairness objective. We show that even approximately optimal optimization of the expected Nash Social Welfare is computationally intractable even in the tabular case. Nevertheless, we provide a novel adaptation of Q-learning that combines nonlinear scalarized learning updates and non-stationary action selection to learn effective policies for optimizing nonlinear welfare functions. We show that our algorithm is provably convergent, and we demonstrate experimentally that our approach outperforms techniques based on linear scalarization, mixtures of optimal linear scalarizations, or stationary action selection for the Nash Social Welfare Objective.

Federated Learning (FL) has emerged as a promising Machine Learning paradigm, enabling multiple users to collaboratively train a shared model while preserving their local data. To minimize computing and communication costs associated with parameter transfer, it is common practice in FL to select a subset of clients in each training round. This selection must consider both system and static heterogeneity. Therefore, we propose FLASH-RL, a framework that utilizes Double Deep QLearning (DDQL) to address both system and static heterogeneity in FL. FLASH-RL introduces a new reputation-based utility function to evaluate client contributions based on their current and past performances. Additionally, an adapted DDQL algorithm is proposed to expedite the learning process. Experimental results on MNIST and CIFAR-10 datasets have shown FLASH-RL's effectiveness in achieving a balanced trade-off between model performance and end-to-end latency against existing solutions. Indeed, FLASH-RL reduces latency by up to 24.83% compared to FedAVG and 24.67% compared to FAVOR. It also reduces the training rounds by up to 60.44% compared to FedAVG and +76% compared to FAVOR. In fall detection using the MobiAct dataset, FLASH-RL outperforms FedAVG by up to 2.82% in model's performance and reduces latency by up to 34.75%. Additionally, FLASH-RL achieves the target performance faster, with up to a 45.32% reduction in training rounds compared to FedAVG.

Radio tagging and tracking are fundamental to understanding the movements and habitats of wildlife in their natural environments. Today, the most widespread, widely applicable technology for gathering data relies on experienced scientists armed with handheld radio telemetry equipment to locate low-power radio transmitters attached to wildlife from the ground. Although aerial robots can transform labor-intensive conservation tasks, the realization of autonomous systems for tackling task complexities under real-world conditions remains a challenge. We developed ConservationBots-- small aerial robots for tracking multiple, dynamic, radio-tagged wildlife. The aerial robot achieves robust localization performance and fast task completion times--significant for energy-limited aerial systems while avoiding close encounters with potential, counter-productive disturbances to wildlife. Our approach overcomes the technical and practical problems posed by combining a lightweight sensor with new concepts: i) planning to determine both trajectory and measurement actions guided by an information-theoretic objective, which allows the robot to strategically select near-instantaneous range-only measurements to achieve faster localization, and time-consuming sensor rotation actions to acquire bearing measurements and achieve robust tracking performance; ii) a bearing detector more robust to noise and iii) a tracking algorithm formulation robust to missed and false detections experienced in real-world conditions. We conducted extensive studies: simulations built upon complex signal propagation over high-resolution elevation data on diverse geographical terrains; field testing; studies with wombats (Lasiorhinus latifrons; nocturnal, vulnerable species dwelling in underground warrens) and tracking comparisons with a highly experienced biologist to validate the effectiveness of our aerial robot and demonstrate the significant advantages over the manual method.

Recent digital advances have popularized predictive maintenance (PMx), offering enhanced efficiency, automation, accuracy, cost savings, and independence in maintenance. Yet, it continues to face numerous limitations such as poor explainability, sample inefficiency of data-driven methods, complexity of physics-based methods, and limited generalizability and scalability of knowledge-based methods. This paper proposes leveraging Digital Twins (DTs) to address these challenges and enable automated PMx adoption at larger scales. While we argue that DTs have this transformative potential, they have not yet reached the level of maturity needed to bridge these gaps in a standardized way. Without a standard definition for such evolution, this transformation lacks a solid foundation upon which to base its development. This paper provides a requirement-based roadmap supporting standardized PMx automation using DT technologies. A systematic approach comprising two primary stages is presented. First, we methodically identify the Informational Requirements (IRs) and Functional Requirements (FRs) for PMx, which serve as a foundation from which any unified framework must emerge. Our approach to defining and using IRs and FRs to form the backbone of any PMx DT is supported by the track record of IRs and FRs being successfully used as blueprints in other areas, such as for product development within the software industry. Second, we conduct a thorough literature review spanning fields to determine the ways in which these IRs and FRs are currently being used within DTs, enabling us to point to the specific areas where further research is warranted to support the progress and maturation of requirement-based PMx DTs.

Graph collaborative filtering, which learns user and item representations through message propagation over the user-item interaction graph, has been shown to effectively enhance recommendation performance. However, most current graph collaborative filtering models mainly construct the interaction graph on a single behavior domain (e.g. click), even though users exhibit various types of behaviors on real-world platforms, including actions like click, cart, and purchase. Furthermore, due to variations in user engagement, there exists an imbalance in the scale of different types of behaviors. For instance, users may click and view multiple items but only make selective purchases from a small subset of them. How to alleviate the behavior imbalance problem and utilize information from the multiple behavior graphs concurrently to improve the target behavior conversion (e.g. purchase) remains underexplored. To this end, we propose IMGCF, a simple but effective model to alleviate behavior data imbalance for multi-behavior graph collaborative filtering. Specifically, IMGCF utilizes a multi-task learning framework for collaborative filtering on multi-behavior graphs. Then, to mitigate the data imbalance issue, IMGCF improves representation learning on the sparse behavior by leveraging representations learned from the behavior domain with abundant data volumes. Experiments on two widely-used multi-behavior datasets demonstrate the effectiveness of IMGCF.