Artificial Intelligence (AI), along with the recent progress in biomedical language understanding, is gradually changing medical practice. With the development of biomedical language understanding benchmarks, AI applications are widely used in the medical field. However, most benchmarks are limited to English, which makes it challenging to replicate many of the successes in English for other languages. To facilitate research in this direction, we collect real-world biomedical data and present the first Chinese Biomedical Language Understanding Evaluation (CBLUE) benchmark: a collection of natural language understanding tasks including named entity recognition, information extraction, clinical diagnosis normalization, single-sentence/sentence-pair classification, and an associated online platform for model evaluation, comparison, and analysis. To establish evaluation on these tasks, we report empirical results with the current 11 pre-trained Chinese models, and experimental results show that state-of-the-art neural models perform by far worse than the human ceiling. Our benchmark is released at \url{https://tianchi.aliyun.com/dataset/dataDetail?dataId=95414&lang=en-us}.

To improve users' experience as they navigate the myriad of options offered by online marketplaces, it is essential to have well-organized product catalogs. One key ingredient to that is the availability of product attributes such as color or material. However, on some marketplaces such as Rakuten-Ichiba, which we focus on, attribute information is often incomplete or even missing. One promising solution to this problem is to rely on deep models pre-trained on large corpora to predict attributes from unstructured data, such as product descriptive texts and images (referred to as modalities in this paper). However, we find that achieving satisfactory performance with this approach is not straightforward but rather the result of several refinements, which we discuss in this paper. We provide a detailed description of our approach to attribute extraction, from investigating strong single-modality methods, to building a solid multimodal model combining textual and visual information. One key component of our multimodal architecture is a novel approach to seamlessly combine modalities, which is inspired by our single-modality investigations. In practice, we notice that this new modality-merging method may suffer from a modality collapse issue, i.e., it neglects one modality. Hence, we further propose a mitigation to this problem based on a principled regularization scheme. Experiments on Rakuten-Ichiba data provide empirical evidence for the benefits of our approach, which has been also successfully deployed to Rakuten-Ichiba. We also report results on publicly available datasets showing that our model is competitive compared to several recent multimodal and unimodal baselines.

Full Bayesian posteriors are rarely analytically tractable, which is why real-world Bayesian inference heavily relies on approximate techniques. Approximations generally differ from the true posterior and require diagnostic tools to assess whether the inference can still be trusted. We investigate a new approach to diagnosing approximate inference: the approximation mismatch is attributed to a change in the inductive bias by treating the approximations as exact and reverse-engineering the corresponding prior. We show that the problem is more complicated than it appears to be at first glance, because the solution generally depends on the observation. By reframing the problem in terms of incompatible conditional distributions we arrive at a natural solution: the Gibbs prior. The resulting diagnostic is based on pseudo-Gibbs sampling, which is widely applicable and easy to implement. We illustrate how the Gibbs prior can be used to discover the inductive bias in a controlled Gaussian setting and for a variety of Bayesian models and approximations.

From speed RADAR guns to high tech riot gear, the role of technology in policing is always evolving. Police departments around the world are always looking for new technologies to help them monitor and protect their communities. One of the areas where police departments always need as much help as possible is in monitoring. Police can't prevent crimes or intervene if they don't know that they're happening, but watching the community is time-intensive work – as much as they may try, they can't be everywhere at once. Two of the most significant advances in this area in recent years has been the wide implementation of CCTV cameras and, more recently, facial recognition software.

There are now some 120,000 warehouses globally, and another 50,000 are likely to be added before 2025. Over the next few years, more robots will be deployed into these warehouses--the logistics market--than in all other application categories combined, including farming, medicine, and home use. Just as the 1960s saw the mechanization of industry, with an accompanying boom in productivity and prosperity, the 2020s will be the dawn of the robotification of services. Industrial robots came into use in 1961 when General Motors (G.M.) installed a simple robotic arm on its New Jersey production line. The machine had been invented by Unimation, a company founded by the father of robotics, Joseph Engelberger--a self-professed Isaac Asimov enthusiast.

We present a simple and effective approach for posterior uncertainty quantification in deep operator networks (DeepONets); an emerging paradigm for supervised learning in function spaces. We adopt a frequentist approach based on randomized prior ensembles, and put forth an efficient vectorized implementation for fast parallel inference on accelerated hardware. Through a collection of representative examples in computational mechanics and climate modeling, we show that the merits of the proposed approach are fourfold. (1) It can provide more robust and accurate predictions when compared against deterministic DeepONets. (2) It shows great capability in providing reliable uncertainty estimates on scarce data-sets with multi-scale function pairs. (3) It can effectively detect out-of-distribution and adversarial examples. (4) It can seamlessly quantify uncertainty due to model bias, as well as noise corruption in the data. Finally, we provide an optimized JAX library called {\em UQDeepONet} that can accommodate large model architectures, large ensemble sizes, as well as large data-sets with excellent parallel performance on accelerated hardware, thereby enabling uncertainty quantification for DeepONets in realistic large-scale applications.

Meta learning aims at learning a model that can quickly adapt to unseen tasks. Widely used meta learning methods include model agnostic meta learning (MAML), implicit MAML, Bayesian MAML. Thanks to its ability of modeling uncertainty, Bayesian MAML often has advantageous empirical performance. However, the theoretical understanding of Bayesian MAML is still limited, especially on questions such as if and when Bayesian MAML has provably better performance than MAML. In this paper, we aim to provide theoretical justifications for Bayesian MAML's advantageous performance by comparing the meta test risks of MAML and Bayesian MAML. In the meta linear regression, under both the distribution agnostic and linear centroid cases, we have established that Bayesian MAML indeed has provably lower meta test risks than MAML. We verify our theoretical results through experiments.

For more than 20 years, the Landsat 7 satellite circled Earth every 99 minutes or so, capturing images of almost all the planet's surface each 16 days. One of many craft that observed the changing globe, it revealed melting glaciers in Greenland, the growth of shrimp farms in Mexico, and the extent of deforestation in Papua New Guinea. But after Landsat 7 ran short on fuel, its useful life effectively ended. In space, regular servicing has not been an option. Now, though, NASA has a potential fix for such enfeebled satellites.

New Zealand is developing an approach to supporting the ethical adoption of AI -- one that is focused on building an AI ecosystem on a foundation of trust, equity and accessibility right from the onset. A crucial part of this approach is to involve key stakeholders in the planning. And that is exactly the reason why the government has designed the system so every New Zealander and every technology expert who matters can contribute. The success of this ITP requires us to form a consensus view on the scope of our ambition and how this can be achieved with actions and initiatives that are sufficiently realistic to bring about meaningful change – both short and longer-term. Wellington published a draft that should jumpstart its pursuit of an ethical AI ecosystem: the Industry Transformation Plan (ITP) which covers its overall digital transformation road map.

Abalone is a common name for any of a group of small to very large sea snails, marine gastropod molluscs in the family Haliotidae. Other common names are ear shells, sea ears, and muttonfish or muttonshells in Australia, ormer in the UK, perlemoen in South Africa, and paua in New Zealand. The age of abalone is determined by cutting the shell through the cone, staining it, and counting the number of rings through a microscope a boring and time consuming task. Other measurements, which are easier to obtain, are used to predict the age. Given is the attribute name, attribute type, the measurement unit and a brief description.