In NLP, models are usually evaluated by reporting single-number performance scores on a number of readily available benchmarks, without much deeper analysis. Here, we argue that - especially given the well-known fact that benchmarks often contain biases, artefacts, and spurious correlations - deeper results analysis should become the de-facto standard when presenting new models or benchmarks. We present a tool that researchers can use to study properties of the dataset and the influence of those properties on their models' behaviour. Our Text Characterization Toolkit includes both an easy-to-use annotation tool, as well as off-the-shelf scripts that can be used for specific analyses. We also present use-cases from three different domains: we use the tool to predict what are difficult examples for given well-known trained models and identify (potentially harmful) biases and heuristics that are present in a dataset.

Autonomous agents that act with each other on behalf of humans are becoming more common in many social domains, such as customer service, transportation, and health care. In such social situations greedy strategies can reduce the positive outcome for all agents, such as leading to stop-and-go traffic on highways, or causing a denial of service on a communications channel. Instead, we desire autonomous decision-making for efficient performance while also considering equitability of the group to avoid these pitfalls. Unfortunately, in complex situations it is far easier to design machine learning objectives for selfish strategies than for equitable behaviors. Here we present a simple way to reward groups of agents in both evolution and reinforcement learning domains by the performance of their weakest member. We show how this yields ``fairer'' more equitable behavior, while also maximizing individual outcomes, and we show the relationship to biological selection mechanisms of group-level selection and inclusive fitness theory.

The COVID-19 crisis has emphasized the need for scientific methods such as machine learning to speed up the discovery of solutions to the pandemic. Harnessing machine learning techniques requires quality data, skilled personnel and advanced compute infrastructure. In Africa, however, machine learning competencies and compute infrastructures are limited. This paper demonstrates a cross-border collaborative capacity building approach to the application of machine learning techniques in discovering answers to COVID-19 questions.

Time-Series Forecasting is a powerful data modeling discipline that analyzes historical observations to predict future values of a time-series. It has been utilized in numerous applications, including but not limited to economics, meteorology, and health. In this paper, we use time-series forecasting techniques to model and predict the future incidence of chickenpox. To achieve this, we implement and simulate multiple models and data preprocessing techniques on a Hungary-collected dataset. We demonstrate that the LSTM model outperforms all other models in the vast majority of the experiments in terms of county-level forecasting, whereas the SARIMAX model performs best at the national level. We also demonstrate that the performance of the traditional data preprocessing method is inferior to that of the data preprocessing method that we have proposed.

Top Two algorithms arose as an adaptation of Thompson sampling to best arm identification in multi-armed bandit models (Russo, 2016), for parametric families of arms. They select the next arm to sample from by randomizing among two candidate arms, a leader and a challenger. Despite their good empirical performance, theoretical guarantees for fixed-confidence best arm identification have only been obtained when the arms are Gaussian with known variances. In this paper, we provide a general analysis of Top Two methods, which identifies desirable properties of the leader, the challenger, and the (possibly non-parametric) distributions of the arms. As a result, we obtain theoretically supported Top Two algorithms for best arm identification with bounded distributions. Our proof method demonstrates in particular that the sampling step used to select the leader inherited from Thompson sampling can be replaced by other choices, like selecting the empirical best arm.

An intelligent control strategy is proposed to eliminate the actuator saturation problem that exists in the trajectory tracking process of unmanned underwater vehicles (UUV). The control strategy consists of two parts: for the kinematic modeling part, a fuzzy logic-refined backstepping control is developed to achieve control velocities within acceptable ranges and errors of small fluctuations; on the basis of the velocities deducted by the improved kinematic control, the sliding mode control (SMC) is introduced in the dynamic modeling to obtain corresponding torques and forces that should be applied to the vehicle body. With the control velocities computed by the kinematic model and applied forces derived by the dynamic model, the robustness and accuracy of the UUV trajectory without actuator saturation can be achieved.

We analyze feature learning in infinite-width neural networks trained with gradient flow through a self-consistent dynamical field theory. We construct a collection of deterministic dynamical order parameters which are inner-product kernels for hidden unit activations and gradients in each layer at pairs of time points, providing a reduced description of network activity through training. These kernel order parameters collectively define the hidden layer activation distribution, the evolution of the neural tangent kernel, and consequently output predictions. We show that the field theory derivation recovers the recursive stochastic process of infinite-width feature learning networks obtained from Yang and Hu (2021) with Tensor Programs . For deep linear networks, these kernels satisfy a set of algebraic matrix equations. For nonlinear networks, we provide an alternating sampling procedure to self-consistently solve for the kernel order parameters. We provide comparisons of the self-consistent solution to various approximation schemes including the static NTK approximation, gradient independence assumption, and leading order perturbation theory, showing that each of these approximations can break down in regimes where general self-consistent solutions still provide an accurate description. Lastly, we provide experiments in more realistic settings which demonstrate that the loss and kernel dynamics of CNNs at fixed feature learning strength is preserved across different widths on a CIFAR classification task.

We introduce Mintaka, a complex, natural, and multilingual dataset designed for experimenting with end-to-end question-answering models. Mintaka is composed of 20,000 question-answer pairs collected in English, annotated with Wikidata entities, and translated into Arabic, French, German, Hindi, Italian, Japanese, Portuguese, and Spanish for a total of 180,000 samples. Mintaka includes 8 types of complex questions, including superlative, intersection, and multi-hop questions, which were naturally elicited from crowd workers. We run baselines over Mintaka, the best of which achieves 38% hits@1 in English and 31% hits@1 multilingually, showing that existing models have room for improvement. We release Mintaka at https://github.com/amazon-research/mintaka.

Few-shot abstractive summarization has become a challenging task in natural language generation. To support it, we designed a novel soft prompts architecture coupled with a prompt pre-training plus fine-tuning paradigm that is effective and tunes only extremely light parameters. The soft prompts include continuous input embeddings across an encoder and a decoder to fit the structure of the generation models. Importantly, a novel inner-prompt placed in the text is introduced to capture document-level information. The aim is to devote attention to understanding the document that better prompts the model to generate document-related content. The first step in the summarization procedure is to conduct prompt pre-training with self-supervised pseudo-data. This teaches the model basic summarizing capabilities. The model is then fine-tuned with few-shot examples. Experimental results on the CNN/DailyMail and XSum datasets show that our method, with only 0.1% of the parameters, outperforms full-model tuning where all model parameters are tuned. It also surpasses Prompt Tuning by a large margin and delivers competitive results against Prefix-Tuning with 3% of the parameters.

Knowledge graph embedding aims to predict the missing relations between entities in knowledge graphs. Tensor-decomposition-based models, such as ComplEx, provide a good trade-off between efficiency and expressiveness, that is crucial because of the large size of real world knowledge graphs. The recent multi-partition embedding interaction (MEI) model subsumes these models by using the block term tensor format and provides a systematic solution for the trade-off. However, MEI has several drawbacks, some of which carried from its subsumed tensor-decomposition-based models. In this paper, we address these drawbacks and introduce the Multi-partition Embedding Interaction iMproved beyond block term format (MEIM) model, with independent core tensor for ensemble effects and soft orthogonality for max-rank mapping, in addition to multi-partition embedding. MEIM improves expressiveness while still being highly efficient, helping it to outperform strong baselines and achieve state-of-the-art results on difficult link prediction benchmarks using fairly small embedding sizes. The source code is released at https://github.com/tranhungnghiep/MEIM-KGE.