近年来，人工智能（AI）的领域已经见证了巨大的增长，然而，持续发展的一些最紧迫的挑战是电子计算机架构所面临的基本带宽，能效和速度限制。利用用于执行神经网络推理操作的光子处理器越来越感兴趣，但是这些网络目前使用标准数字电子培训。这里，我们提出了由CMOS兼容的硅光子架构实现的神经网络的片上训练，以利用大规模平行，高效和快速数据操作的电位。我们的方案采用直接反馈对准训练算法，它使用错误反馈而不是错误反向化而培训神经网络，并且可以在每秒乘以数万亿乘以量的速度运行，同时每次MAC操作消耗小于一个微微约会。光子架构利用并行化矩阵 - 向量乘法利用微址谐振器阵列，用于沿着单个波导总线处理多通道模拟信号，以便原位计算每个神经网络层的梯度向量，这是在后向通过期间执行的最昂贵的操作。 。我们还通过片上MAC操作结果实验地示意使用MNIST数据集进行培训深度神经网络。我们的高效，超快速神经网络训练的新方法展示了光子学作为执行AI应用的有希望的平台。

语言模型既展示了定量的改进，又展示了新的定性功能，随着规模的增加。尽管它们具有潜在的变革性影响，但这些新能力的特征却很差。为了为未来的研究提供信息，为破坏性的新模型能力做准备，并改善社会有害的效果，至关重要的是，我们必须了解目前和近乎未来的能力和语言模型的局限性。为了应对这一挑战，我们介绍了超越模仿游戏基准（Big Bench）。 Big Bench目前由204个任务组成，由132家机构的442位作者贡献。任务主题是多样的，从语言学，儿童发展，数学，常识性推理，生物学，物理学，社会偏见，软件开发等等。 Big-Bench专注于被认为超出当前语言模型的功能的任务。我们评估了OpenAI的GPT型号，Google内部密集变压器体系结构和大型基础上的开关稀疏变压器的行为，跨越了数百万到数十亿个参数。此外，一个人类专家评估者团队执行了所有任务，以提供强大的基准。研究结果包括：模型性能和校准都随规模改善，但绝对的术语（以及与评估者的性能相比）；在模型类中的性能非常相似，尽管带有稀疏性。逐渐和预测的任务通常涉及大量知识或记忆成分，而在临界规模上表现出“突破性”行为的任务通常涉及多个步骤或组成部分或脆性指标；社交偏见通常会随着含糊不清的环境而随着规模而增加，但这可以通过提示来改善。

背景：12个引线ECG是心血管疾病的核心诊断工具。在这里，我们描述并分析了一个集成的深度神经网络架构，从12个引导eCG分类了24个心脏异常。方法：我们提出了挤压和激发reset，以自动学习来自12个引主ECG的深度特征，以识别24个心脏病。在最终完全连接的层中，随着年龄和性别特征增强了深度特征。使用约束网格搜索设置每个类的输出阈值。为了确定为什么该模型的预测不正确，两个专家诊所人员独立地解释了一组关于左轴偏差的一次无序的ECG。结果：采用定制加权精度度量，我们达到了0.684的5倍交叉验证得分，灵敏度和特异性分别为0.758和0.969。我们在完整的测试数据中得分0.520，并在官方挑战排名中排名第21中。在一系列被错误分类的心电图中，两个临床医生和训练标签之间的协议差（临床医生1：Kappa = -0.057，临床医生2：Kappa = -0.159）。相比之下，临床医生之间的协议非常高（Kappa = 0.92）。讨论：与在相同数据上培训的模型相比，所提出的预测模型很好地对验证和隐藏的测试数据进行了良好。我们还发现培训标签的相当不一致，这可能会阻碍更准确的模型的开发。

天文学家通常已经着手通过从头开始创建自己的表示来解决监督的机器学习问题。我们表明，经过训练的深度学习模型，可以回答每个星系动物园贴花问题问题，即学习星系的有意义的语义表示，这些语义表示对于从未训练过的新任务很有用。我们利用这些表示形式优于最近对研究大型星系样本至关重要的实际任务的方法。第一个任务是识别与查询星系相似的形态的星系。给定一个星系为人类分配了一个免费文本标签（例如“ #diffuse”），我们可以找到与大多数标签匹配该标签的星系。第二个任务是确定特定研究人员最有趣的异常。我们的方法在识别最有趣的100个异常（由Galaxy Zoo 2志愿者判断）方面是100％准确的。第三个任务是调整模型来仅使用少数新标记的星系解决新任务。与从陆地图像（ImageNet）或从头开始训练的模型相比，从我们的表示形式进行微调的模型可以更好地识别环形星系。我们用很少的新标签解决每个任务；一个（用于相似性搜索）或数百个（用于异常检测或微调）。这挑战了长期以来的观点，即深度监督方法需要新的大型标签数据集，以便在天文学中实际使用。为了帮助社区受益于我们验证的模型，我们发布了我们的微调代码Zoobot。没有先前经验的研究人员可以访问Zoobot。

In this paper, we propose a novel technique, namely INVALIDATOR, to automatically assess the correctness of APR-generated patches via semantic and syntactic reasoning. INVALIDATOR reasons about program semantic via program invariants while it also captures program syntax via language semantic learned from large code corpus using the pre-trained language model. Given a buggy program and the developer-patched program, INVALIDATOR infers likely invariants on both programs. Then, INVALIDATOR determines that a APR-generated patch overfits if: (1) it violates correct specifications or (2) maintains errors behaviors of the original buggy program. In case our approach fails to determine an overfitting patch based on invariants, INVALIDATOR utilizes a trained model from labeled patches to assess patch correctness based on program syntax. The benefit of INVALIDATOR is three-fold. First, INVALIDATOR is able to leverage both semantic and syntactic reasoning to enhance its discriminant capability. Second, INVALIDATOR does not require new test cases to be generated but instead only relies on the current test suite and uses invariant inference to generalize the behaviors of a program. Third, INVALIDATOR is fully automated. We have conducted our experiments on a dataset of 885 patches generated on real-world programs in Defects4J. Experiment results show that INVALIDATOR correctly classified 79% overfitting patches, accounting for 23% more overfitting patches being detected by the best baseline. INVALIDATOR also substantially outperforms the best baselines by 14% and 19% in terms of Accuracy and F-Measure, respectively.

When robots learn reward functions using high capacity models that take raw state directly as input, they need to both learn a representation for what matters in the task -- the task ``features" -- as well as how to combine these features into a single objective. If they try to do both at once from input designed to teach the full reward function, it is easy to end up with a representation that contains spurious correlations in the data, which fails to generalize to new settings. Instead, our ultimate goal is to enable robots to identify and isolate the causal features that people actually care about and use when they represent states and behavior. Our idea is that we can tune into this representation by asking users what behaviors they consider similar: behaviors will be similar if the features that matter are similar, even if low-level behavior is different; conversely, behaviors will be different if even one of the features that matter differs. This, in turn, is what enables the robot to disambiguate between what needs to go into the representation versus what is spurious, as well as what aspects of behavior can be compressed together versus not. The notion of learning representations based on similarity has a nice parallel in contrastive learning, a self-supervised representation learning technique that maps visually similar data points to similar embeddings, where similarity is defined by a designer through data augmentation heuristics. By contrast, in order to learn the representations that people use, so we can learn their preferences and objectives, we use their definition of similarity. In simulation as well as in a user study, we show that learning through such similarity queries leads to representations that, while far from perfect, are indeed more generalizable than self-supervised and task-input alternatives.

The latent space of autoencoders has been improved for clustering image data by jointly learning a t-distributed embedding with a clustering algorithm inspired by the neighborhood embedding concept proposed for data visualization. However, multivariate tabular data pose different challenges in representation learning than image data, where traditional machine learning is often superior to deep tabular data learning. In this paper, we address the challenges of learning tabular data in contrast to image data and present a novel Gaussian Cluster Embedding in Autoencoder Latent Space (G-CEALS) algorithm by replacing t-distributions with multivariate Gaussian clusters. Unlike current methods, the proposed approach independently defines the Gaussian embedding and the target cluster distribution to accommodate any clustering algorithm in representation learning. A trained G-CEALS model extracts a quality embedding for unseen test data. Based on the embedding clustering accuracy, the average rank of the proposed G-CEALS method is 1.4 (0.7), which is superior to all eight baseline clustering and cluster embedding methods on seven tabular data sets. This paper shows one of the first algorithms to jointly learn embedding and clustering to improve multivariate tabular data representation in downstream clustering.

An unbiased scene graph generation (SGG) algorithm referred to as Skew Class-balanced Re-weighting (SCR) is proposed for considering the unbiased predicate prediction caused by the long-tailed distribution. The prior works focus mainly on alleviating the deteriorating performances of the minority predicate predictions, showing drastic dropping recall scores, i.e., losing the majority predicate performances. It has not yet correctly analyzed the trade-off between majority and minority predicate performances in the limited SGG datasets. In this paper, to alleviate the issue, the Skew Class-balanced Re-weighting (SCR) loss function is considered for the unbiased SGG models. Leveraged by the skewness of biased predicate predictions, the SCR estimates the target predicate weight coefficient and then re-weights more to the biased predicates for better trading-off between the majority predicates and the minority ones. Extensive experiments conducted on the standard Visual Genome dataset and Open Image V4 \& V6 show the performances and generality of the SCR with the traditional SGG models.

In this paper we discuss the theory used in the design of an open source lightmorphic signatures analysis toolkit (LSAT). In addition to providing a core functionality, the software package enables specific optimizations with its modular and customizable design. To promote its usage and inspire future contributions, LSAT is publicly available. By using a self-supervised neural network and augmented machine learning algorithms, LSAT provides an easy-to-use interface with ample documentation. The experiments demonstrate that LSAT improves the otherwise tedious and error-prone tasks of translating lightmorphic associated data into usable spectrograms, enhanced with parameter tuning and performance analysis. With the provided mathematical functions, LSAT validates the nonlinearity encountered in the data conversion process while ensuring suitability of the forecasting algorithms.

Detecting abrupt changes in data distribution is one of the most significant tasks in streaming data analysis. Although many unsupervised Change-Point Detection (CPD) methods have been proposed recently to identify those changes, they still suffer from missing subtle changes, poor scalability, or/and sensitive to noise points. To meet these challenges, we are the first to generalise the CPD problem as a special case of the Change-Interval Detection (CID) problem. Then we propose a CID method, named iCID, based on a recent Isolation Distributional Kernel (IDK). iCID identifies the change interval if there is a high dissimilarity score between two non-homogeneous temporal adjacent intervals. The data-dependent property and finite feature map of IDK enabled iCID to efficiently identify various types of change points in data streams with the tolerance of noise points. Moreover, the proposed online and offline versions of iCID have the ability to optimise key parameter settings. The effectiveness and efficiency of iCID have been systematically verified on both synthetic and real-world datasets.