预先训练的上下文化文本表示模型学习自然语言的有效表示，以使IT机器可以理解。在注意机制的突破之后，已经提出了新一代预磨模的模型，以便自变压器引入以来实现了良好的性能。来自变压器（BERT）的双向编码器表示已成为语言理解的最先进的模型。尽管取得了成功，但大多数可用的型号已经在印度欧洲语言中培训，但是对代表性的语言和方言的类似研究仍然稀疏。在本文中，我们调查了培训基于单语言变换器的语言模型的可行性，以获得代表语言的特定重点是突尼斯方言。我们评估了我们的语言模型对情感分析任务，方言识别任务和阅读理解问答任务。我们表明使用嘈杂的Web爬网数据而不是结构化数据（维基百科，文章等）更方便这些非标准化语言。此外，结果表明，相对小的Web爬网数据集导致与使用较大数据集获得的那些表现相同的性能。最后，我们在所有三个下游任务中达到或改善了最先进的Tunbert模型。我们释放出Tunbert净化模型和用于微调的数据集。

The Internet of Senses (IoS) holds the promise of flawless telepresence-style communication for all human `receptors' and therefore blurs the difference of virtual and real environments. We commence by highlighting the compelling use cases empowered by the IoS and also the key network requirements. We then elaborate on how the emerging semantic communications and Artificial Intelligence (AI)/Machine Learning (ML) paradigms along with 6G technologies may satisfy the requirements of IoS use cases. On one hand, semantic communications can be applied for extracting meaningful and significant information and hence efficiently exploit the resources and for harnessing a priori information at the receiver to satisfy IoS requirements. On the other hand, AI/ML facilitates frugal network resource management by making use of the enormous amount of data generated in IoS edge nodes and devices, as well as by optimizing the IoS performance via intelligent agents. However, the intelligent agents deployed at the edge are not completely aware of each others' decisions and the environments of each other, hence they operate in a partially rather than fully observable environment. Therefore, we present a case study of Partially Observable Markov Decision Processes (POMDP) for improving the User Equipment (UE) throughput and energy consumption, as they are imperative for IoS use cases, using Reinforcement Learning for astutely activating and deactivating the component carriers in carrier aggregation. Finally, we outline the challenges and open issues of IoS implementations and employing semantic communications, edge intelligence as well as learning under partial observability in the IoS context.

Recent work has demonstrated that natural language processing techniques can support consumer protection by automatically detecting unfair clauses in the Terms of Service (ToS) Agreement. This work demonstrates that transformer-based ToS analysis systems are vulnerable to adversarial attacks. We conduct experiments attacking an unfair-clause detector with universal adversarial triggers. Experiments show that a minor perturbation of the text can considerably reduce the detection performance. Moreover, to measure the detectability of the triggers, we conduct a detailed human evaluation study by collecting both answer accuracy and response time from the participants. The results show that the naturalness of the triggers remains key to tricking readers.

This technical report presents GPS++, the first-place solution to the Open Graph Benchmark Large-Scale Challenge (OGB-LSC 2022) for the PCQM4Mv2 molecular property prediction task. Our approach implements several key principles from the prior literature. At its core our GPS++ method is a hybrid MPNN/Transformer model that incorporates 3D atom positions and an auxiliary denoising task. The effectiveness of GPS++ is demonstrated by achieving 0.0719 mean absolute error on the independent test-challenge PCQM4Mv2 split. Thanks to Graphcore IPU acceleration, GPS++ scales to deep architectures (16 layers), training at 3 minutes per epoch, and large ensemble (112 models), completing the final predictions in 1 hour 32 minutes, well under the 4 hour inference budget allocated. Our implementation is publicly available at: https://github.com/graphcore/ogb-lsc-pcqm4mv2.

The demonstrated success of transfer learning has popularized approaches that involve pretraining models from massive data sources and subsequent finetuning towards a specific task. While such approaches have become the norm in fields such as natural language processing, implementation and evaluation of transfer learning approaches for chemistry are in the early stages. In this work, we demonstrate finetuning for downstream tasks on a graph neural network (GNN) trained over a molecular database containing 2.7 million water clusters. The use of Graphcore IPUs as an AI accelerator for training molecular GNNs reduces training time from a reported 2.7 days on 0.5M clusters to 1.2 hours on 2.7M clusters. Finetuning the pretrained model for downstream tasks of molecular dynamics and transfer to a different potential energy surface took only 8.3 hours and 28 minutes, respectively, on a single GPU.

Angluin的L*算法使用会员资格和等价查询了解了常规语言的最低（完整）确定性有限自动机（DFA）。它的概率近似正确（PAC）版本用足够大的随机会员查询替换等效查询，以使答案获得高级信心。因此，它可以应用于任何类型的（也是非规范）设备，可以将其视为合成自动机的算法，该算法根据观测值抽象该设备的行为。在这里，我们对Angluin的PAC学习算法对通过引入一些噪音从DFA获得的设备感兴趣。更确切地说，我们研究盎格鲁因算法是否会降低噪声并产生与原始设备更接近原始设备的DFA。我们提出了几种介绍噪声的方法：（1）嘈杂的设备将单词的分类W.R.T.倒置。具有很小概率的DFA，（2）嘈杂的设备在询问其分类W.R.T.之前用小概率修改了单词的字母。 DFA和（3）嘈杂的设备结合了W.R.T.单词的分类。 DFA及其分类W.R.T.柜台自动机。我们的实验是在数百个DFA上进行的。直言不讳地表明，我们的主要贡献表明：（1）每当随机过程产生嘈杂的设备时，盎格鲁因算法的行为都很好，（2）但使用结构化的噪声却很差，并且（3）几乎肯定是随机性的产量具有非竞争性语言的系统。

多词表达式（MWE）是一系列单词，共同提出的含义不是从其单个单词中得出的。处理MWE的任务在许多自然语言处理（NLP）应用中至关重要，包括机器翻译和术语提取。因此，在不同领域中检测MWE是一个重要的研究主题。在本文中，我们探索了最新的神经变压器，以检测花和植物名称中的MWES。我们在由植物和花朵百科全书创建的数据集上评估了不同的变压器模型。我们从经验上表明，Transformer模型模型优于基于长期记忆（LSTM）的先前神经模型。

在处理机器学习模型（例如图形神经网络（GNN））中的一批图表时，通常将几个小图组合到一个整体图中以加速处理并减少填充的开销。例如，这是PYG库中支持的。但是，小图的尺寸对于节点和边缘的数量可能会有很大的变化，因此，组合图的大小仍然可能有很大差异，尤其是对于小批量大小而言。因此，仍然产生过多的填充和浪费计算的成本。本文提出了一种新方法 - 元组包装 - 用于生成导致最小开销的批次。该算法扩展了最近引入的序列填料方法，以在（| nodes |，| edges |）的2D元组上工作。单调启发式词被应用于元组值的2D直方图，以定义填充直方图箱的优先级，以及目标以达到节点数量和边缘数量的限制。实验验证了多个数据集上算法的有效性。

通常，用于训练排名模型的数据受到标签噪声。例如，在Web搜索中，由于ClickStream数据创建的标签是嘈杂的，这是因为诸如SERP上的项目描述中的信息不足，用户查询重新进行的，以及不稳定的或意外的用户行为。在实践中，很难处理标签噪声而不对标签生成过程做出强烈的假设。结果，如果不考虑标签噪声，从业人员通常会直接在此嘈杂的数据上训练他们的学习到秩（LTR）模型。令人惊讶的是，我们经常看到以这种方式训练的LTR模型的出色表现。在这项工作中，我们描述了一类耐噪声的LTR损失，即使在类条件标签噪声的背景下，经验风险最小化也是一致的程序。我们还开发了常用损失函数的耐噪声类似物。实验结果进一步支持了我们理论发现的实际意义。

自动驾驶汽车是一项不断发展的技术，旨在通过自动操作从车道变更到超车来提高安全性，可访问性，效率和便利性。超车是自动驾驶汽车最具挑战性的操作之一，当前的自动超车技术仅限于简单情况。本文研究了如何通过允许动作流产来提高自主超车的安全性。我们提出了一个基于深层Q网络的决策过程，以确定是否以及何时需要中止超车的操作。拟议的算法在与交通情况不同的模拟中进行了经验评估，这表明所提出的方法可以改善超车手动过程中的安全性。此外，使用自动班车Iseauto在现实世界实验中证明了该方法。