Recent advances in federated learning have demonstrated its promising capability to learn on decentralized datasets. However, a considerable amount of work has raised concerns due to the potential risks of adversaries participating in the framework to poison the global model for an adversarial purpose. This paper investigates the feasibility of model poisoning for backdoor attacks through rare word embeddings of NLP models. In text classification, less than 1% of adversary clients suffices to manipulate the model output without any drop in the performance on clean sentences. For a less complex dataset, a mere 0.1% of adversary clients is enough to poison the global model effectively. We also propose a technique specialized in the federated learning scheme called Gradient Ensemble, which enhances the backdoor performance in all our experimental settings.

Artificial Intelligence (AI) is a potentially transformational technology that will have broad social, economic, national security, and geopolitical implications for the United States and the world.1 AI is not one particular technology but a general-purpose technology combining software and hardware in systems that enable technologies (machine learning, knowledge representation, and other forms of computerized approximation of human intelligence). This general-purpose nature means that AI could have wide-ranging economic impacts across manufacturing, transportation, health, education, and many other sectors. In 2018, the McKinsey Global Institute estimated that AI could add around 16 percent, or $13 trillion, to global output by 2030.2 Since then COVID-19 has further accelerated the use of AI. While the United States is the world leader in AI, China is catching up fast (and may lead in some areas) and other governments are expanding their own AI capacity. Rather than a zero-sum game, many such efforts can be additive, benefiting global welfare. The U.S. can encourage and support AI efforts that seek to develop and compete on fair terms. Other national policies--China's above all--seek to erect barriers to free and open development of AI, appropriating the benefits for their national champions and applying AI as a geopolitical lever. Such policies could distort the development and benefits of AI for humanity, make the world less secure for the U.S. and allies, and markets less receptive to U.S. products and services. To foster AI policies that support development of beneficial, trustworthy, and robust artificial intelligence will require international engagement by the United States and cooperation among like-minded democracies that are leaders in artificial intelligence.

Scientists revealed the use of artificial intelligence to predict the sugarcane industry's performance efficiently.

Though model robustness has been extensively studied in language understanding, the robustness of Seq2Seq generation remains understudied. In this paper, we conduct the first quantitative analysis on the robustness of pre-trained Seq2Seq models. We find that even current SOTA pre-trained Seq2Seq model (BART) is still vulnerable, which leads to significant degeneration in faithfulness and informativeness for text generation tasks. This motivated us to further propose a novel adversarial augmentation framework, namely AdvSeq, for generally improving faithfulness and informativeness of Seq2Seq models via enhancing their robustness. AdvSeq automatically constructs two types of adversarial augmentations during training, including implicit adversarial samples by perturbing word representations and explicit adversarial samples by word swapping, both of which effectively improve Seq2Seq robustness. Extensive experiments on three popular text generation tasks demonstrate that AdvSeq significantly improves both the faithfulness and informativeness of Seq2Seq generation under both automatic and human evaluation settings.

Fact-checking requires retrieving evidence related to a claim under investigation. The task can be formulated as question generation based on a claim, followed by question answering. However, recent question generation approaches assume that the answer is known and typically contained in a passage given as input, whereas such passages are what is being sought when verifying a claim. In this paper, we present {\it Varifocal}, a method that generates questions based on different focal points within a given claim, i.e.\ different spans of the claim and its metadata, such as its source and date. Our method outperforms previous work on a fact-checking question generation dataset on a wide range of automatic evaluation metrics. These results are corroborated by our manual evaluation, which indicates that our method generates more relevant and informative questions. We further demonstrate the potential of focal points in generating sets of clarification questions for product descriptions.

Combining the visual modality with pretrained language models has been surprisingly effective for simple descriptive tasks such as image captioning. More general text generation however remains elusive. We take a step back and ask: How do these models work for more complex generative tasks, i.e. conditioning on both text and images? Are multimodal models simply visually adapted language models, or do they combine they reason jointly over modalities? We investigate these questions in the context of self-rationalization (jointly generating task labels/answers and free-text explanations) of three tasks: (i) visual question answering in VQA-X, (ii) visual commonsense reasoning in VCR, and (iii) visual-textual entailment in e-SNLI-VE. We show that recent unimodal advances, CLIP image representations and scaling of language models, do not consistently improve self-rationalization in multimodal tasks. We find that no single model type works universally best across tasks, datasets, and finetuning data sizes. Our findings motivate the need for novel general backbones approach that move text generation from images and text beyond image captioning.

Tokenisation is the first step in almost all NLP tasks, and state-of-the-art transformer-based language models all use subword tokenisation algorithms to process input text. Existing algorithms have problems, often producing tokenisations of limited linguistic validity, and representing equivalent strings differently depending on their position within a word. We hypothesise that these problems hinder the ability of transformer-based models to handle complex words, and suggest that these problems are a result of allowing tokens to include spaces. We thus experiment with an alternative tokenisation approach where spaces are always treated as individual tokens. Specifically, we apply this modification to the BPE and Unigram algorithms. We find that our modified algorithms lead to improved performance on downstream NLP tasks that involve handling complex words, whilst having no detrimental effect on performance in general natural language understanding tasks. Intrinsically, we find our modified algorithms give more morphologically correct tokenisations, in particular when handling prefixes. Given the results of our experiments, we advocate for always treating spaces as individual tokens as an improved tokenisation method.

An eight-year-old boy with cerebral palsy has been restricted to a wheelchair all his life, but he recently took his first steps thanks to a robotic exoskeleton designed specifically for children. David Zabala was fitted with a new Atlas 2030 exoskeleton that features mechanical joints that adapt to his motions, allowing him to finally walk freely. The innovation is part of a therapy method being used at a facility in Mexico City, because research shows that allowing paralyzed children the opportunity to walk'not only extends their life expectancy and enhances their physical well-being, but also improves their self-esteem.' The suit helps'to achieve in record time rehabilitation goals' that would take months to achieve with conventional therapies, Guadalupe Maldonado, director of Mexico's Association for People with Cerebral Palsy told AFP. The suit was create by Elena García Armada who won the 2022 European Inventor Award for the innovation.

New video footage emerged this week of a four-legged robot goalkeeper making some impressive saves in a research lab. It's part an exciting scientific field to develop machines that can play sports as well as – or even better than – human professionals. Here, MailOnline takes a look at the robots taking the world of sports by storm, from the six-legged skiier to the ping pong ace and the expert curler. The robot dog was trained using'reinforcement learning' or RL – a subset of machine learning. RL allows an AI-driven system (sometimes referred to as an agent) to learn through trial and error using feedback from its actions. This feedback is either negative or positive, signalled as punishment or reward with, of course, the aim of maximising the reward function.

We propose a novel open-domain question answering (ODQA) framework for answering single/multi-hop questions across heterogeneous knowledge sources. The key novelty of our method is the introduction of the intermediary modules into the current retriever-reader pipeline. Unlike previous methods that solely rely on the retriever for gathering all evidence in isolation, our intermediary performs a chain of reasoning over the retrieved set. Specifically, our method links the retrieved evidence with its related global context into graphs and organizes them into a candidate list of evidence chains. Built upon pretrained language models, our system achieves competitive performance on two ODQA datasets, OTT-QA and NQ, against tables and passages from Wikipedia. In particular, our model substantially outperforms the previous state-of-the-art on OTT-QA with an exact match score of 47.3 (45 % relative gain).