Constructing benchmarks that test the abilities of modern natural language understanding models is difficult - pre-trained language models exploit artifacts in benchmarks to achieve human parity, but still fail on adversarial examples and make errors that demonstrate a lack of common sense. In this work, we propose gamification as a framework for data construction. The goal of players in the game is to compose questions that mislead a rival AI while using specific phrases for extra points. The game environment leads to enhanced user engagement and simultaneously gives the game designer control over the collected data, allowing us to collect high-quality data at scale. Using our method we create CommonsenseQA 2.0, which includes 14,343 yes/no questions, and demonstrate its difficulty for models that are orders-of-magnitude larger than the AI used in the game itself. Our best baseline, the T5-based Unicorn with 11B parameters achieves an accuracy of 70.2%, substantially higher than GPT-3 (52.9%) in a few-shot inference setup. Both score well below human performance which is at 94.1%.

The uprising of Artificial Intelligence machines (hereinafter referred as "AI") is a popular and intriguing subject for many science fiction works. The advancement of AI machines and their progression with respect to playing a significant role in our lives has increased exponentially in the past few years. The future possibilities of this technology has stirred a hornets' nest of innumerable possibilities. As we witness AI machines overlapping with Intellectual Property Rights (IPR), it gives rise to many questions concerning legal discipline. When the earliest substantial work in the field of Artificial Intelligence was concluded in the mid-20th century by the British logician and computer pioneer, Alan Mathison Turing, nobody could have imagined that there will be an attempt towards an assimilation of technical solutions created by an AI machines into the scope of patent law.

Terrorism is a major problem worldwide, causing thousands of fatalities and billions of dollars in damage every year. Toward the end of better understanding and mitigating these attacks, we present a set of machine learning models that learn from localized news data in order to predict whether a terrorist attack will occur on a given calendar date and in a given state. The best model--a Random Forest that learns from a novel variable-length moving average representation of the feature space--achieves area under the receiver operating characteristic scores $> .667$ on four of the five states that were impacted most by terrorism between 2015 and 2018. Our key findings include that modeling terrorism as a set of independent events, rather than as a continuous process, is a fruitful approach--especially when the events are sparse and dissimilar. Additionally, our results highlight the need for localized models that account for differences between locations. From a machine learning perspective, we found that the Random Forest model outperformed several deep models on our multimodal, noisy, and imbalanced data set, thus demonstrating the efficacy of our novel feature representation method in such a context. We also show that its predictions are relatively robust to time gaps between attacks and observed characteristics of the attacks. Finally, we analyze factors that limit model performance, which include a noisy feature space and small amount of available data. These contributions provide an important foundation for the use of machine learning in efforts against terrorism in the United States and beyond.

The world of Artificial intelligence is not a new concept for researchers. This technology has been discovered and is constantly being developed for an unimaginably long period. There are still some myths found that give hints of Mechanical men in Ancient Greek and Egypt. The answer to a frequently asked question of who invented Artificial intelligence is very broad, as many people contributed to the invention and timely success. It goes back to the classic philosophers who attempted to portray human thinking as a symbolic system.

As Amazon's CTO since 2005, Werner Vogels has observed macro trends around the technology industry, giving him a unique perspective to distinguish substantive progress from mere fads. He recently published his views on what he sees in store in 2022 for cloud technology and the technology world in general. In his post, Werner lays out five core predictions, regarding the growth of artificial intelligence (AI), the abundance of data, the power of machine learning (ML), architecting for sustainability, and the full reach of connectivity via the Internet, backed by cloud-based resources. I think his analysis provides takeaways for startups overall and where they might seek to create value in the next year. Let's dive into each of these anticipated developments and their potential impact on the world of startups.

When you "work through" a problem or issue that requires a decision, you likely feel as if you're going through a linear checklist. But that's not how the human brain operates; it processes in a non-linear pattern. And this is essentially how deep learning, a subset of artificial intelligence (AI), works too. Deep learning, at its essence, learns from examples -- the way the human brain does. It's imitating the way humans acquire certain types of knowledge.

Relying on random matrix theory (RMT), this paper studies asymmetric order-$d$ spiked tensor models with Gaussian noise. Using the variational definition of the singular vectors and values of (Lim, 2005), we show that the analysis of the considered model boils down to the analysis of an equivalent spiked symmetric block-wise random matrix, that is constructed from contractions of the studied tensor with the singular vectors associated to its best rank-1 approximation. Our approach allows the exact characterization of the almost sure asymptotic singular value and alignments of the corresponding singular vectors with the true spike components, when $\frac{n_i}{\sum_{j=1}^d n_j}\to c_i\in [0, 1]$ with $n_i$'s the tensor dimensions. In contrast to other works that rely mostly on tools from statistical physics to study random tensors, our results rely solely on classical RMT tools such as Stein's lemma. Finally, classical RMT results concerning spiked random matrices are recovered as a particular case.

We introduce an algorithm for computing geodesics on sampled manifolds that relies on simulation of quantum dynamics on a graph embedding of the sampled data. Our approach exploits classic results in semiclassical analysis and the quantum-classical correspondence, and forms a basis for techniques to learn the manifold from which a dataset is sampled, and subsequently for nonlinear dimensionality reduction of high-dimensional datasets. We illustrate the new algorithm with data sampled from model manifolds and also by a clustering demonstration based on COVID-19 mobility data. Finally, our method reveals interesting connections between the discretization provided by data sampling and quantization.

Existing explanation methods for black-box supervised learning models generally work by building local models that explain the models behaviour for a particular data item. It is possible to make global explanations, but the explanations may have low fidelity for complex models. Most of the prior work on explainable models has been focused on classification problems, with less attention on regression. We propose a new manifold visualization method, SLISEMAP, that at the same time finds local explanations for all of the data items and builds a two-dimensional visualization of model space such that the data items explained by the same model are projected nearby. We provide an open source implementation of our methods, implemented by using GPU-optimized PyTorch library. SLISEMAP works both on classification and regression models. We compare SLISEMAP to most popular dimensionality reduction methods and some local explanation methods. We provide mathematical derivation of our problem and show that SLISEMAP provides fast and stable visualizations that can be used to explain and understand black box regression and classification models.

The rapid emergence of airborne platforms and imaging sensors are enabling new forms of aerial surveillance due to their unprecedented advantages in scale, mobility, deployment and covert observation capabilities. This paper provides a comprehensive overview of human-centric aerial surveillance tasks from a computer vision and pattern recognition perspective. It aims to provide readers with an in-depth systematic review and technical analysis of the current state of aerial surveillance tasks using drones, UAVs and other airborne platforms. The main object of interest is humans, where single or multiple subjects are to be detected, identified, tracked, re-identified and have their behavior analyzed. More specifically, for each of these four tasks, we first discuss unique challenges in performing these tasks in an aerial setting compared to a ground-based setting. We then review and analyze the aerial datasets publicly available for each task, and delve deep into the approaches in the aerial literature and investigate how they presently address the aerial challenges. We conclude the paper with discussion on the missing gaps and open research questions to inform future research avenues.