Out-of-distribution (OoD) generalization occurs when representation learning encounters a distribution shift. This occurs frequently in practice when training and testing data come from different environments. Covariate shift is a type of distribution shift that occurs only in the input data, while the concept distribution stays invariant. We propose RIA - Regularization for Invariance with Adversarial training, a new method for OoD generalization under convariate shift. Motivated by an analogy to $Q$-learning, it performs an adversarial exploration for counterfactual data environments. These new environments are induced by adversarial label invariant data augmentations that prevent a collapse to an in-distribution trained learner. It works with many existing OoD generalization methods for covariate shift that can be formulated as constrained optimization problems. We develop an alternating gradient descent-ascent algorithm to solve the problem in the context of causally generated graph data, and perform extensive experiments on OoD graph classification for various kinds of synthetic and natural distribution shifts. We demonstrate that our method can achieve high accuracy compared with OoD baselines.

Popular post-training pruning methods such as Wanda and RIA are known for their simple, yet effective, designs that have shown exceptional empirical performance. Wanda optimizes performance through calibrated activations during pruning, while RIA emphasizes the relative, rather than absolute, importance of weight elements. Despite their practical success, a thorough theoretical foundation explaining these outcomes has been lacking. This paper introduces new theoretical insights that redefine the standard minimization objective for pruning, offering a deeper understanding of the factors contributing to their success. Our study extends beyond these insights by proposing complementary strategies that consider both input activations and weight significance. We validate these approaches through rigorous experiments, demonstrating substantial enhancements over existing methods. Furthermore, we introduce a novel training-free fine-tuning approach $R^2$-DSnoT that incorporates relative weight importance and a regularized decision boundary within a dynamic pruning-and-growing framework, significantly outperforming strong baselines and establishing a new state of the art.

Significant strides have been made in natural language tasks, largely attributed to the emergence of powerful large language models (LLMs). These models, pre-trained on extensive and diverse corpora, have become increasingly capable of comprehending the intricacies of language. Despite the abundance of LLMs for many high-resource languages, the availability of such models remains limited for European Portuguese. We introduce Gl\'orIA, a robust European Portuguese decoder LLM. To pre-train Gl\'orIA, we assembled a comprehensive PT-PT text corpus comprising 35 billion tokens from various sources. We present our pre-training methodology, followed by an assessment of the model's effectiveness on multiple downstream tasks. Additionally, to evaluate our models' language modeling capabilities, we introduce CALAME-PT (Context-Aware LAnguage Modeling Evaluation for Portuguese), the first Portuguese zero-shot language-modeling benchmark. Evaluation shows that Gl\'orIA significantly outperforms existing open PT decoder models in language modeling and that it can generate sound, knowledge-rich, and coherent PT-PT text. The model also exhibits strong potential for various downstream tasks.

Deep learning models have demonstrated remarkable capabilities in learning complex patterns and concepts from training data. However, recent findings indicate that these models tend to rely heavily on simple and easily discernible features present in the background of images rather than the main concepts or objects they are intended to classify. This phenomenon poses a challenge to image classifiers as the crucial elements of interest in images may be overshadowed. In this paper, we propose a novel approach to address this issue and improve the learning of main concepts by image classifiers. Our central idea revolves around concurrently guiding the model's attention toward the foreground during the classification task. By emphasizing the foreground, which encapsulates the primary objects of interest, we aim to shift the focus of the model away from the dominant influence of the background. To accomplish this, we introduce a mechanism that encourages the model to allocate sufficient attention to the foreground. We investigate various strategies, including modifying the loss function or incorporating additional architectural components, to enable the classifier to effectively capture the primary concept within an image. Additionally, we explore the impact of different foreground attention mechanisms on model performance and provide insights into their effectiveness. Through extensive experimentation on benchmark datasets, we demonstrate the efficacy of our proposed approach in improving the classification accuracy of image classifiers. Our findings highlight the importance of foreground attention in enhancing model understanding and representation of the main concepts within images. The results of this study contribute to advancing the field of image classification and provide valuable insights for developing more robust and accurate deep-learning models.

In "The Code Detectives," two middle school girls who love coding use artificial intelligence to solve mysteries. For 17-year-old author Ria Dosha, writing the book series is a way to advocate for increasing diversity within the technology field. "I've brought a diverse cast of characters to life, with the series centering around Ramona Diaz, a powerful young girl of color," says Ria, a student at Cupertino's Monta Vista High School. "The book series gives young girls strong, fictional role models in technology and AI, and introduces them to AI topics in a compelling way, clearing common misconceptions." Ria writes what shoe knows, and vice versa.

For all of human history, people have spied on one another. To find out what others are doing or planning to do, people have surveilled, monitored, and eavesdropped--using tools that constantly improved but never displaced their human masters. Artificial intelligence (AI) and autonomous systems are changing all of that. In the future, machines will spy on machines in order to know what other machines are doing or are planning to do. Intelligence work will still consist of stealing and protecting secrets, but how those secrets are collected, analyzed, and disseminated will be fundamentally different.

The Trump administration plans to reinterpret the deal between 34 nations in order to allow U.S. defence contractors to sell more American-made drones to various nations, three defence industry executives and a U.S. official told Reuters.

Robots took on a record number of jobs in North American firms last year, the Robotic Industries Association (RIA) said Thursday. According to the RIA's data, 35,880 robots were shipped in 2018 to the U.S., Canada and Mexico, up 7 percent from the previous year. Of those shipments, 16,702 were to non-automotive companies -- a year-on-year increase of 41 percent. The consumer goods sector purchased almost 50 percent more robots in 2018 than in 2017, while life sciences saw an increase of a third. However, shipments to the automotive industry slowed by 12 percent.

Artificial intelligence can help businesses identify new clients and streamline back office operations, but so far, only the largest companies have found the data-heavy tech practical. Early adopters of artificial intelligence are expecting substantial increases in profitability in 2018, according to the Nationwide's Advisor Authority survey of 1,700 advisors. But, those that used big data were also more likely to have sizeable assets under management, over $250 million in AUM and incomes of more than $500,000. That's because, for more modest firms, fewer clients mean less reliable data. "Wealth managers are still struggling to get a consistent line of sight on customer data and that includes things like demographics and transaction data," says William Trout, senior analyst at the consulting firm Celent.

In 2016, after four years of running a health and fitness app that lets customers track their daily food and workout routines, health-tech startup HealthifyMe found that it was sitting on millions of data points about its users' lifestyle habits. The Bengaluru-based company's customers were using the app to not only log and track their health regimes, but also talk to nutritionists or fitness coaches. This meant that HealthifyMe had data on everything from users' food and workout logs to the questions they asked the nutritionists and the responses they received. So the company, which is backed by IDG Ventures India, Inventus Capital, and Blume Ventures, among others, decided to feed all this information to a machine-learning algorithm and help fitness coaches respond better to app users. That project has now developed into a customer-facing programme where an AI-driven bot talks to over 25,000 of HealthifyMe's paid subscribers, similar to how Google Assistant or Amazon's Alexa operate.