Showing papers by "Carnegie Mellon University published in 2021"

OpenPose: Realtime Multi-Person 2D Pose Estimation Using Part Affinity Fields

Abstract: Realtime multi-person 2D pose estimation is a key component in enabling machines to have an understanding of people in images and videos. In this work, we present a realtime approach to detect the 2D pose of multiple people in an image. The proposed method uses a nonparametric representation, which we refer to as Part Affinity Fields (PAFs), to learn to associate body parts with individuals in the image. This bottom-up system achieves high accuracy and realtime performance, regardless of the number of people in the image. In previous work, PAFs and body part location estimation were refined simultaneously across training stages. We demonstrate that a PAF-only refinement rather than both PAF and body part location refinement results in a substantial increase in both runtime performance and accuracy. We also present the first combined body and foot keypoint detector, based on an internal annotated foot dataset that we have publicly released. We show that the combined detector not only reduces the inference time compared to running them sequentially, but also maintains the accuracy of each component individually. This work has culminated in the release of OpenPose, the first open-source realtime system for multi-person 2D pose detection, including body, foot, hand, and facial keypoints.

Advances and open problems in federated learning

Abstract: The term Federated Learning was coined as recently as 2016 to describe a machine learning setting where multiple entities collaborate in solving a machine learning problem, under the coordination of a central server or service provider. Each client’s raw data is stored locally and not exchanged or transferred; instead, focused updates intended for immediate aggregation are used to achieve the learning objective. Since then, the topic has gathered much interest across many different disciplines and the realization that solving many of these interdisciplinary problems likely requires not just machine learning but techniques from distributed optimization, cryptography, security, differential privacy, fairness, compressed sensing, systems, information theory, statistics, and more. This monograph has contributions from leading experts across the disciplines, who describe the latest state-of-the art from their perspective. These contributions have been carefully curated into a comprehensive treatment that enables the reader to understand the work that has been done and get pointers to where effort is required to solve many of the problems before Federated Learning can become a reality in practical systems. Researchers working in the area of distributed systems will find this monograph an enlightening read that may inspire them to work on the many challenging issues that are outlined. This monograph will get the reader up to speed quickly and easily on what is likely to become an increasingly important topic: Federated Learning.

An Empirical Study of Training Self-Supervised Vision Transformers

Abstract: This paper does not describe a novel method Instead, it studies a straightforward, incremental, yet must-know baseline given the recent progress in computer vision: self-supervised learning for Vision Transformers (ViT) While the training recipes for standard convolutional networks have been highly mature and robust, the recipes for ViT are yet to be built, especially in the self-supervised scenarios where training becomes more challenging In this work, we go back to basics and investigate the effects of several fundamental components for training self-supervised ViT We observe that instability is a major issue that degrades accuracy, and it can be hidden by apparently good results We reveal that these results are indeed partial failure, and they can be improved when training is made more stable We benchmark ViT results in MoCo v3 and several other self-supervised frameworks, with ablations in various aspects We discuss the currently positive evidence as well as challenges and open questions We hope that this work will provide useful data points and experience for future research

mRNA vaccines for infectious diseases: principles, delivery and clinical translation.

Abstract: Over the past several decades, messenger RNA (mRNA) vaccines have progressed from a scepticism-inducing idea to clinical reality. In 2020, the COVID-19 pandemic catalysed the most rapid vaccine development in history, with mRNA vaccines at the forefront of those efforts. Although it is now clear that mRNA vaccines can rapidly and safely protect patients from infectious disease, additional research is required to optimize mRNA design, intracellular delivery and applications beyond SARS-CoV-2 prophylaxis. In this Review, we describe the technologies that underlie mRNA vaccines, with an emphasis on lipid nanoparticles and other non-viral delivery vehicles. We also overview the pipeline of mRNA vaccines against various infectious disease pathogens and discuss key questions for the future application of this breakthrough vaccine platform.

Lifestyle and mental health disruptions during COVID-19.

Abstract: Using a longitudinal dataset linking biometric and survey data from several cohorts of young adults before and during the COVID-19 pandemic ([Formula: see text]), we document large disruptions to physical activity, sleep, time use, and mental health. At the onset of the pandemic, average steps decline from 10,000 to 4,600 steps per day, sleep increases by 25 to 30 min per night, time spent socializing declines by over half to less than 30 min, and screen time more than doubles to over 5 h per day. Over the course of the pandemic from March to July 2020 the proportion of participants at risk for clinical depression ranges from 46% to 61%, up to a 90% increase in depression rates compared to the same population just prior to the pandemic. Our analyses suggest that disruption to physical activity is a leading risk factor for depression during the pandemic. However, restoration of those habits through a short-term intervention does not meaningfully improve mental well-being.

AWAC: Accelerating Online Reinforcement Learning with Offline Datasets

Abstract: Reinforcement learning provides an appealing formalism for learning control policies from experience. However, the classic active formulation of reinforcement learning necessitates a lengthy active exploration process for each behavior, making it difficult to apply in real-world settings. If we can instead allow reinforcement learning to effectively use previously collected data to aid the online learning process, where the data could be expert demonstrations or more generally any prior experience, we could make reinforcement learning a substantially more practical tool. While a number of recent methods have sought to learn offline from previously collected data, it remains exceptionally difficult to train a policy with offline data and improve it further with online reinforcement learning. In this paper we systematically analyze why this problem is so challenging, and propose an algorithm that combines sample-efficient dynamic programming with maximum likelihood policy updates, providing a simple and effective framework that is able to leverage large amounts of offline data and then quickly perform online fine-tuning of reinforcement learning policies. We show that our method enables rapid learning of skills with a combination of prior demonstration data and online experience across a suite of difficult dexterous manipulation and benchmark tasks.

Toward Active-Site Tailoring in Heterogeneous Catalysis by Atomically Precise Metal Nanoclusters with Crystallographic Structures.

Abstract: Heterogeneous catalysis involves solid-state catalysts, among which metal nanoparticles occupy an important position. Unfortunately, no two nanoparticles from conventional synthesis are the same at the atomic level, though such regular nanoparticles can be highly uniform at the nanometer level (e.g., size distribution ∼5%). In the long pursuit of well-defined nanocatalysts, a recent success is the synthesis of atomically precise metal nanoclusters protected by ligands in the size range from tens to hundreds of metal atoms (equivalently 1-3 nm in core diameter). More importantly, such nanoclusters have been crystallographically characterized, just like the protein structures in enzyme catalysis. Such atomically precise metal nanoclusters merge the features of well-defined homogeneous catalysts (e.g., ligand-protected metal centers) and enzymes (e.g., protein-encapsulated metal clusters of a few atoms bridged by ligands). The well-defined nanoclusters with their total structures available constitute a new class of model catalysts and hold great promise in fundamental catalysis research, including the atomically precise size dependent activity, control of catalytic selectivity by metal structure and surface ligands, structure-property relationships at the atomic-level, insights into molecular activation and catalytic mechanisms, and the identification of active sites on nanocatalysts. This Review summarizes the progress in the utilization of atomically precise metal nanoclusters for catalysis. These nanocluster-based model catalysts have enabled heterogeneous catalysis research at the single-atom and single-electron levels. Future efforts are expected to achieve more exciting progress in fundamental understanding of the catalytic mechanisms, the tailoring of active sites at the atomic level, and the design of new catalysts with high selectivity and activity under mild conditions.

HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units

Abstract: Self-supervised approaches for speech representation learning are challenged by three unique problems: (1) there are multiple sound units in each input utterance, (2) there is no lexicon of input sound units during the pre-training phase, and (3) sound units have variable lengths with no explicit segmentation. To deal with these three problems, we propose the Hidden-Unit BERT (HuBERT) approach for self-supervised speech representation learning, which utilizes an offline clustering step to provide aligned target labels for a BERT-like prediction loss. A key ingredient of our approach is applying the prediction loss over the masked regions only, which forces the model to learn a combined acoustic and language model over the continuous inputs. HuBERT relies primarily on the consistency of the unsupervised clustering step rather than the intrinsic quality of the assigned cluster labels. Starting with a simple k-means teacher of 100 clusters, and using two iterations of clustering, the HuBERT model either matches or improves upon the state-of-the-art wav2vec 2.0 performance on the Librispeech (960 h) and Libri-light (60,000 h) benchmarks with 10 min, 1 h, 10 h, 100 h, and 960 h fine-tuning subsets. Using a 1B parameter model, HuBERT shows up to 19% and 13% relative WER reduction on the more challenging dev-other and test-other evaluation subsets. 1 2

A Comprehensive Survey on Graph Anomaly Detection with Deep Learning

Abstract: Over the last forty years, researches on anomalies have received intensified interests and the burst of information has attracted more attention on anomalies because of their significance in a wide range of disciplines. Anomaly detection, which aims to identify these rare observations, is among the most vital tasks and has shown its power in preventing detrimental events, such as financial fraud, network intrusion, and social spam, from happening. The detection task is typically solved by detecting outlying data in the features space and inherently overlooks the structural information. Graphs have been prevalently used to preserve structural information, and this raises the graph anomaly detection problem - identifying anomalous graph objects (nodes, edges, sub-graphs, and graphs). However, conventional anomaly detection techniques cannot well solve this problem because of the complexity of graph data. For the aptitudes of deep learning in breaking these limitations, graph anomaly detection with deep learning has received intensified studies recently. In this survey, we aim to provide a systematic and comprehensive review of the contemporary deep learning techniques for graph anomaly detection. We also highlight twelve extensive future research directions according to our survey results covering emerging problems introduced by graph data, anomaly detection and real applications.

Scene Text Detection and Recognition: The Deep Learning Era

Abstract: With the rise and development of deep learning, computer vision has been tremendously transformed and reshaped. As an important research area in computer vision, scene text detection and recognition has been inevitably influenced by this wave of revolution, consequentially entering the era of deep learning. In recent years, the community has witnessed substantial advancements in mindset, methodology and performance. This survey is aimed at summarizing and analyzing the major changes and significant progresses of scene text detection and recognition in the deep learning era. Through this article, we devote to: (1) introduce new insights and ideas; (2) highlight recent techniques and benchmarks; (3) look ahead into future trends. Specifically, we will emphasize the dramatic differences brought by deep learning and remaining grand challenges. We expect that this review paper would serve as a reference book for researchers in this field. Related resources are also collected in our Github repository ( https://github.com/Jyouhou/SceneTextPapers ).

Handbook of Model Checking

Abstract: Model checking is a computer-assisted method for the analysis of dynamical systems that can be modeled by state-transition systems. Drawing from research traditions in mathematical logic, programming languages, hardware design, and theoretical computer science, model checking is now widely used for the verification of hardware and software in industry. The editors and authors of this handbook are among the world's leading researchers in this domain, and the 32 contributed chapters present a thorough view of the origin, theory, and application of model checking. In particular, the editors classify the advances in this domain and the chapters of the handbook in terms of two recurrent themes that have driven much of the research agenda: the algorithmic challenge, that is, designing model-checking algorithms that scale to real-life problems; and the modeling challenge, that is, extending the formalism beyond Kripke structures and temporal logic. The book will be valuable for researchers and graduate students engaged with the development of formal methods and verification tools.

Accurate Determination of the Neutron Skin Thickness of ^{208}Pb through Parity-Violation in Electron Scattering.

Abstract: We report a precision measurement of the parity-violating asymmetry A_{PV} in the elastic scattering of longitudinally polarized electrons from ^{208}Pb. We measure A_{PV}=550±16(stat)±8(syst) parts per billion, leading to an extraction of the neutral weak form factor F_{W}(Q^{2}=0.00616 GeV^{2})=0.368±0.013. Combined with our previous measurement, the extracted neutron skin thickness is R_{n}-R_{p}=0.283±0.071 fm. The result also yields the first significant direct measurement of the interior weak density of ^{208}Pb: ρ_{W}^{0}=-0.0796±0.0036(exp)±0.0013(theo) fm^{-3} leading to the interior baryon density ρ_{b}^{0}=0.1480±0.0036(exp)±0.0013(theo) fm^{-3}. The measurement accurately constrains the density dependence of the symmetry energy of nuclear matter near saturation density, with implications for the size and composition of neutron stars.

Algorithms for Verifying Deep Neural Networks

Abstract: Neural networks have been widely used in many applications, such as image classification and understanding, language processing, and control of autonomous systems. These networks work by mapping inputs to outputs through a sequence of layers. At each layer, the input to that layer undergoes an affine transformation followed by a simple nonlinear transformation before being passed to the next layer. Neural networks are being used for increasingly important tasks, and in some cases, incorrect outputs can lead to costly consequences, hence validation of correctness at each layer is vital. The sheer size of the networks makes this not feasible using traditional methods. In this monograph, the authors survey a class of methods that are capable of formally verifying properties of deep neural networks. In doing so, they introduce a unified mathematical framework for verifying neural networks, classify existing methods under this framework, provide pedagogical implementations of existing methods, and compare those methods on a set of benchmark problems. Algorithms for Verifying Deep Neural Networks serves as a tutorial for students and professionals interested in this emerging field as well as a benchmark to facilitate the design of new verification algorithms.

An electrically conductive silver–polyacrylamide–alginate hydrogel composite for soft electronics

Abstract: Hydrogels offer tissue-like compliance, stretchability, fracture toughness, ionic conductivity and compatibility with biological tissues. However, their electrical conductivity ( 350 S cm−1) and is capable of delivering direct current while maintaining soft compliance (Young’s modulus < 10 kPa) and deformability. Micrometre-sized silver flakes are suspended in a polyacrylamide–alginate hydrogel matrix and, after going through a partial dehydration process, the flakes form percolating networks that are electrically conductive and robust to mechanical deformations. To illustrate the capabilities of our silver–hydrogel composite, we use the material in a stingray-inspired swimmer and a neuromuscular electrical stimulation electrode. A hydrogel composite that consists of micrometre-sized silver flakes suspended in a polyacrylamide–alginate hydrogel matrix exhibits a high electrical conductivity of over 350 S cm−1 and a low Young’s modulus of less than 10 kPa.

Dynabench: Rethinking Benchmarking in NLP.

Abstract: We introduce Dynabench, an open-source platform for dynamic dataset creation and model benchmarking. Dynabench runs in a web browser and supports human-and-model-in-the-loop dataset creation: annotators seek to create examples that a target model will misclassify, but that another person will not. In this paper, we argue that Dynabench addresses a critical need in our community: contemporary models quickly achieve outstanding performance on benchmark tasks but nonetheless fail on simple challenge examples and falter in real-world scenarios. With Dynabench, dataset creation, model development, and model assessment can directly inform each other, leading to more robust and informative benchmarks. We report on four initial NLP tasks, illustrating these concepts and highlighting the promise of the platform, and address potential objections to dynamic benchmarking as a new standard for the field.

Advances in neural rendering

Abstract: Loss functions for Neural Rendering Jun-Yan Zhu

Self-Supervised Deep Correlation Tracking

Abstract: The training of a feature extraction network typically requires abundant manually annotated training samples, making this a time-consuming and costly process. Accordingly, we propose an effective self-supervised learning-based tracker in a deep correlation framework (named: self-SDCT). Motivated by the forward-backward tracking consistency of a robust tracker, we propose a multi-cycle consistency loss as self-supervised information for learning feature extraction network from adjacent video frames. At the training stage, we generate pseudo-labels of consecutive video frames by forward-backward prediction under a Siamese correlation tracking framework and utilize the proposed multi-cycle consistency loss to learn a feature extraction network. Furthermore, we propose a similarity dropout strategy to enable some low-quality training sample pairs to be dropped and also adopt a cycle trajectory consistency loss in each sample pair to improve the training loss function. At the tracking stage, we employ the pre-trained feature extraction network to extract features and utilize a Siamese correlation tracking framework to locate the target using forward tracking alone. Extensive experimental results indicate that the proposed self-supervised deep correlation tracker (self-SDCT) achieves competitive tracking performance contrasted to state-of-the-art supervised and unsupervised tracking methods on standard evaluation benchmarks.

A Comprehensive Survey of Neural Architecture Search: Challenges and Solutions

Abstract: Deep learning has made substantial breakthroughs in many fields due to its powerful automatic representation capabilities. It has been proven that neural architecture design is crucial to the feature representation of data and the final performance. However, the design of the neural architecture heavily relies on the researchers’ prior knowledge and experience. And due to the limitations of humans’ inherent knowledge, it is difficult for people to jump out of their original thinking paradigm and design an optimal model. Therefore, an intuitive idea would be to reduce human intervention as much as possible and let the algorithm automatically design the neural architecture. Neural Architecture Search (NAS) is just such a revolutionary algorithm, and the related research work is complicated and rich. Therefore, a comprehensive and systematic survey on the NAS is essential. Previously related surveys have begun to classify existing work mainly based on the key components of NAS: search space, search strategy, and evaluation strategy. While this classification method is more intuitive, it is difficult for readers to grasp the challenges and the landmark work involved. Therefore, in this survey, we provide a new perspective: beginning with an overview of the characteristics of the earliest NAS algorithms, summarizing the problems in these early NAS algorithms, and then providing solutions for subsequent related research work. In addition, we conduct a detailed and comprehensive analysis, comparison, and summary of these works. Finally, we provide some possible future research directions.

Client Selection in Federated Learning: Convergence Analysis and Power-of-Choice Selection Strategies

Abstract: Federated learning is a distributed optimization paradigm that enables a large number of resource-limited client nodes to cooperatively train a model without data sharing. Several works have analyzed the convergence of federated learning by accounting of data heterogeneity, communication and computation limitations, and partial client participation. However, they assume unbiased client participation, where clients are selected at random or in proportion of their data sizes. In this paper, we present the first convergence analysis of federated optimization for biased client selection strategies, and quantify how the selection bias affects convergence speed. We reveal that biasing client selection towards clients with higher local loss achieves faster error convergence. Using this insight, we propose Power-of-Choice, a communication- and computation-efficient client selection framework that can flexibly span the trade-off between convergence speed and solution bias. We also propose an extension of Power-of-Choice that is able to maintain convergence speed improvement while diminishing the selection skew. Our experiments demonstrate that Power-of-Choice strategies converge up to 3 × faster and give 10% higher test accuracy than the baseline random selection.

Constraints on Dark Matter Properties from Observations of Milky Way Satellite Galaxies

Abstract: We perform a comprehensive study of Milky Way (MW) satellite galaxies to constrain the fundamental properties of dark matter (DM). This analysis fully incorporates inhomogeneities in the spatial distribution and detectability of MW satellites and marginalizes over uncertainties in the mapping between galaxies and DM halos, the properties of the MW system, and the disruption of subhalos by the MW disk. Our results are consistent with the cold, collisionless DM paradigm and yield the strongest cosmological constraints to date on particle models of warm, interacting, and fuzzy dark matter. At 95% confidence, we report limits on (i) the mass of thermal relic warm DM, mWDM>6.5 keV (free-streaming length, λfs≲10h-1 kpc), (ii) the velocity-independent DM-proton scattering cross section, σ0 2.9×10-21 eV (de Broglie wavelength, λdB≲0.5 kpc). These constraints are complementary to other observational and laboratory constraints on DM properties.

Optimized Graph Convolution Recurrent Neural Network for Traffic Prediction

Abstract: Traffic prediction is a core problem in the intelligent transportation system and has broad applications in the transportation management and planning, and the main challenge of this field is how to efficiently explore the spatial and temporal information of traffic data. Recently, various deep learning methods, such as convolution neural network (CNN), have shown promising performance in traffic prediction. However, it samples traffic data in regular grids as the input of CNN, thus it destroys the spatial structure of the road network. In this paper, we introduce a graph network and propose an optimized graph convolution recurrent neural network for traffic prediction, in which the spatial information of the road network is represented as a graph. Additionally, distinguishing with most current methods using a simple and empirical spatial graph, the proposed method learns an optimized graph through a data-driven way in the training phase, which reveals the latent relationship among the road segments from the traffic data. Lastly, the proposed method is evaluated on three real-world case studies, and the experimental results show that the proposed method outperforms state-of-the-art traffic prediction methods.

RMA: Rapid Motor Adaptation for Legged Robots

Abstract: Successful real-world deployment of legged robots would require them to adapt in real-time to unseen scenarios like changing terrains, changing payloads, wear and tear. This paper presents Rapid Motor Adaptation (RMA) algorithm to solve this problem of real-time online adaptation in quadruped robots. RMA consists of two components: a base policy and an adaptation module. The combination of these components enables the robot to adapt to novel situations in fractions of a second. RMA is trained completely in simulation without using any domain knowledge like reference trajectories or predefined foot trajectory generators and is deployed on the A1 robot without any fine-tuning. We train RMA on a varied terrain generator using bioenergetics-inspired rewards and deploy it on a variety of difficult terrains including rocky, slippery, deformable surfaces in environments with grass, long vegetation, concrete, pebbles, stairs, sand, etc. RMA shows state-of-the-art performance across diverse real-world as well as simulation experiments. Project Webpage and Videos: https://ashish-kmr.github.io/rma-legged-robots/

Best practices in machine learning for chemistry.

Abstract: Statistical tools based on machine learning are becoming integrated into chemistry research workflows. We discuss the elements necessary to train reliable, repeatable and reproducible models, and recommend a set of guidelines for machine learning reports.

GSum: A General Framework for Guided Neural Abstractive Summarization

Abstract: Neural abstractive summarization models are flexible and can produce coherent summaries, but they are sometimes unfaithful and can be difficult to control. While previous studies attempt to provide different types of guidance to control the output and increase faithfulness, it is not clear how these strategies compare and contrast to each other. In this paper, we propose a general and extensible guided summarization framework (GSum) that can effectively take different kinds of external guidance as input, and we perform experiments across several different varieties. Experiments demonstrate that this model is effective, achieving state-of-the-art performance according to ROUGE on 4 popular summarization datasets when using highlighted sentences as guidance. In addition, we show that our guided model can generate more faithful summaries and demonstrate how different types of guidance generate qualitatively different summaries, lending a degree of controllability to the learned models.

Correlated insulating states at fractional fillings of the WS2/WSe2 moiré lattice

Abstract: The strong electron interactions in the minibands formed in moire superlattices of van der Waals materials, such as twisted graphene and transition metal dichalcogenides, make such systems a fascinating platform with which to study strongly correlated states1–19. In most systems, the correlated states appear when the moire lattice is filled by an integer number of electrons per moire unit cell. Recently, correlated states at fractional fillings of 1/3 and 2/3 holes per moire unit cell have been reported in the WS2/WSe2 hetero-bilayer, hinting at the long-range nature of the electron interaction16. Here we observe a series of correlated insulating states at fractional fillings of the moire minibands on both electron- and hole-doped sides in angle-aligned WS2/WSe2 hetero-bilayers, with certain states persisting at temperatures up to 120 K. Simulations reveal that these insulating states correspond to ordering of electrons in the moire lattice with a periodicity much larger than the moire unit cell, indicating a surprisingly strong and long-range interaction beyond the nearest neighbours. Twisted bilayers of WS2 and WSe2 have correlated states that correspond to real-space ordering of the electrons on a length scale much longer than the moire pattern.

Sensing of COVID-19 Antibodies in Seconds via Aerosol Jet Nanoprinted Reduced-Graphene-Oxide-Coated 3D Electrodes.

Abstract: Rapid diagnosis is critical for the treatment and prevention of diseases. An advanced nanomaterial-based biosensing platform that detects COVID-19 antibodies within seconds is reported. The biosensing platform is created by 3D nanoprinting of three-dimensional electrodes, coating the electrodes by nanoflakes of reduced-graphene-oxide (rGO), and immobilizing specific viral antigens on the rGO nanoflakes. The electrode is then integrated with a microfluidic device and used in a standard electrochemical cell. When antibodies are introduced on the electrode surface, they selectively bind with the antigens, changing the impedance of the electrical circuit which is detected via impedance spectroscopy. Antibodies to SARS-CoV-2 spike S1 protein and its receptor-binding-domain (RBD) are detected at a limit-of-detection of 2.8 × 10-15 and 16.9 × 10-15 m, respectively, and read by a smartphone-based user interface. The sensor can be regenerated within a minute by introducing a low-pH chemistry that elutes the antibodies from the antigens, allowing successive sensing of test samples using the same sensor. Sensing of S1 and RBD antibodies is specific, which cross-reacts neither with other antibodies such as RBD, S1, and nucleocapsid antibody nor with proteins such as interleukin-6. The proposed sensing platform could also be useful to detect biomarkers for other infectious agents such as Ebola, HIV, and Zika.

Behavioural nudges increase COVID-19 vaccinations.

Abstract: Enhancing vaccine uptake is a critical public health challenge1. Overcoming vaccine hesitancy2,3 and failure to follow through on vaccination intentions3 requires effective communication strategies3,4. Here we present two sequential randomized controlled trials to test the effect of behavioural interventions on the uptake of COVID-19 vaccines. We designed text-based reminders that make vaccination salient and easy, and delivered them to participants drawn from a healthcare system one day (first randomized controlled trial) (n = 93,354 participants; clinicaltrials number NCT04800965) and eight days (second randomized controlled trial) (n = 67,092 individuals; clinicaltrials number NCT04801524) after they received a notification of vaccine eligibility. The first reminder boosted appointment and vaccination rates within the healthcare system by 6.07 (84%) and 3.57 (26%) percentage points, respectively; the second reminder increased those outcomes by 1.65 and 1.06 percentage points, respectively. The first reminder had a greater effect when it was designed to make participants feel ownership of the vaccine dose. However, we found no evidence that combining the first reminder with a video-based information intervention designed to address vaccine hesitancy heightened its effect. We performed online studies (n = 3,181 participants) to examine vaccination intentions, which revealed patterns that diverged from those of the first randomized controlled trial; this underscores the importance of pilot-testing interventions in the field. Our findings inform the design of behavioural nudges for promoting health decisions5, and highlight the value of making vaccination easy and inducing feelings of ownership over vaccines. Two randomized controlled trials demonstrate the ability of text-based behavioural ‘nudges’ to improve the uptake of COVID-19 vaccines, especially when designed to make participants feel ownership over their vaccine dose.

One channel to rule them all? Constraining the origins of binary black holes using multiple formation pathways

Abstract: The second LIGO–Virgo catalog of gravitational-wave (GW) transients has more than quadrupled the observational sample of binary black holes. We analyze this catalog using a suite of five state-of-the-art binary black hole population models covering a range of isolated and dynamical formation channels and infer branching fractions between channels as well as constraints on uncertain physical processes that impact the observational properties of mergers. Given our set of formation models, we find significant differences between the branching fractions of the underlying and detectable populations, and the diversity of detections suggests that multiple formation channels are at play. A mixture of channels is strongly preferred over any single channel dominating the detected population: an individual channel does not contribute to more than sime70% of the observational sample of binary black holes. We calculate the preference between the natal spin assumptions and common-envelope efficiencies in our models, favoring natal spins of isolated black holes of lesssim0.1 and marginally preferring common-envelope efficiencies of gsim2.0 while strongly disfavoring highly inefficient common envelopes. We show that it is essential to consider multiple channels when interpreting GW catalogs, as inference on branching fractions and physical prescriptions becomes biased when contributing formation scenarios are not considered or incorrect physical prescriptions are assumed. Although our quantitative results can be affected by uncertain assumptions in model predictions, our methodology is capable of including models with updated theoretical considerations and additional formation channels.

Understanding Factuality in Abstractive Summarization with FRANK: A Benchmark for Factuality Metrics.

Abstract: Modern summarization models generate highly fluent but often factually unreliable outputs. This motivated a surge of metrics attempting to measure the factuality of automatically generated summaries. Due to the lack of common benchmarks, these metrics cannot be compared. Moreover, all these methods treat factuality as a binary concept and fail to provide deeper insights on the kinds of inconsistencies made by different systems. To address these limitations, we devise a typology of factual errors and use it to collect human annotations of generated summaries from state-of-the-art summarization systems for the CNN/DM and XSum datasets. Through these annotations we identify the proportion of different categories of factual errors and benchmark factuality metrics, showing their correlation with human judgement as well as their specific strengths and weaknesses.