搜索结果：Works

Bengali literature has a rich history of hundreds of years with luminary figures such as Rabindranath Tagore and Kazi Nazrul Islam. However, analytical works involving the most recent advancements in NLP have barely scratched the surface utilizing the enormous volume of the collected works from the writers of the language. In order to bring attention to the analytical study involving the works of Bengali writers and spearhead the text generation endeavours in the style of existing literature, we are introducing RabindraNet, a character level RNN model with stacked-LSTM layers trained on the works of Rabindranath Tagore to produce literary works in his style for multiple genres. We created an extensive dataset as well by compiling the digitized works of Rabindranath Tagore from authentic online sources and published as open source dataset on data science platform Kaggle.

Knowledge distillation (KD) is a core component in the training and deployment of modern generative models, particularly large language models (LLMs). While its empirical benefits are well documented -- enabling smaller student models to emulate the performance of much larger teachers -- the underlying mechanisms by which KD improves generative quality remain poorly understood. In this work, we present a minimal working explanation of KD in generative modeling. Using a controlled simulation with mixtures of Gaussians, we demonstrate that distillation induces a trade-off between precision and recall in the student model. As the teacher distribution becomes more selective, the student concentrates more probability mass on high-likelihood regions at the expense of coverage, which is a behavior modulated by a single entropy-controlling parameter. We then validate this effect in a large-scale language modeling setup using the SmolLM2 family of models. Empirical results reveal the same precision-recall dynamics observed in simulation, where precision corresponds to sample quality and recall to distributional coverage. This precision-recall trade-off in LLMs is found to be especially bene

Extending knowledge by identifying and investigating valuable research questions and problems is a core function of research. Research publications often suggest avenues for future work to extend and build upon their results. Considering these suggestions can contribute to developing research ideas that build upon previous work and produce results that tie into existing knowledge. Usable security and privacy researchers commonly add future work statements to their publications. However, our community lacks an in-depth understanding of their prevalence, quality, and impact on future research. Our work aims to address this gap in the research literature. We reviewed all 27 papers from the 2019 SOUPS proceedings and analyzed their future work statements. Additionally, we analyzed 978 publications that cite any paper from SOUPS 2019 proceedings to assess their future work statements' impact. We find that most papers from the SOUPS 2019 proceedings include future work statements. However, they are often unspecific or ambiguous, and not always easy to find. Therefore, the citing publications often matched the future work statements' content thematically, but rarely explicitly acknowledge

Future works in scientific articles are valuable for researchers and they can guide researchers to new research directions or ideas. In this paper, we mine the future works in scientific articles in order to 1) provide an insight for future work analysis and 2) facilitate researchers to search and browse future works in a research area. First, we study the problem of future work extraction and propose a regular expression based method to address the problem. Second, we define four different categories for the future works by observing the data and investigate the multi-class future work classification problem. Third, we apply the extraction method and the classification model to a paper dataset in the computer science field and conduct a further analysis of the future works. Finally, we design a prototype system to search and demonstrate the future works mined from the scientific papers. Our evaluation results show that our extraction method can get high precision and recall values and our classification model can also get good results and it outperforms several baseline models. Further analysis of the future work sentences also indicates interesting results.

The Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI) has been heavily supporting Machine Learning and Deep Learning research from its foundation in 2012. We have asked six leading ICRI-CI Deep Learning researchers to address the challenge of "Why & When Deep Learning works", with the goal of looking inside Deep Learning, providing insights on how deep networks function, and uncovering key observations on their expressiveness, limitations, and potential. The output of this challenge resulted in five papers that address different facets of deep learning. These different facets include a high-level understating of why and when deep networks work (and do not work), the impact of geometry on the expressiveness of deep networks, and making deep networks interpretable.

This is an overview of some of the works of Avi Wigderson, 2021 Abel prize laureate. Wigderson's contributions span many fields of computer science and mathematics. In this survey we focus on four subfields: cryptography, pseudorandomness, computational complexity lower bounds, and the theory of optimization over symmetric manifolds. Even within those fields, we are not able to mention all of Wigderson's results, let alone cover them in full detail. However, we attempt to give a broad view of each field, as well as describe how Wigderson's papers have answered central questions, made key definitions, forged unexpected connections, or otherwise made lasting changes to our ways of thinking in that field.

This work proposes an overview of the recent semi-supervised learning approaches and related works. Despite the remarkable success of neural networks in various applications, there exist a few formidable constraints, including the need for a large amount of labeled data. Therefore, semi-supervised learning, which is a learning scheme in which scarce labels and a larger amount of unlabeled data are utilized to train models (e.g., deep neural networks), is getting more important. Based on the key assumptions of semi-supervised learning, which are the manifold assumption, cluster assumption, and continuity assumption, the work reviews the recent semi-supervised learning approaches. In particular, the methods in regard to using deep neural networks in a semi-supervised learning setting are primarily discussed. In addition, the existing works are first classified based on the underlying idea and explained, then the holistic approaches that unify the aforementioned ideas are detailed.

The physics laboratory-works creating and operating computer simulations experience is described. A significant amount of laboratory works can be classified as a "black box". The studied physical phenomenon is hidden from direct observation, the control is carried out by means of electrical measuring devices. It is difficult to distinguish physical reality from its imitation when performing such work, so the virtualization of this one does not require realistic images. The schematic representation of the laboratory installation greatly simplifies the process of creating a simulator. A unique set of installation parameters is formed for each student performing laboratory work on the simulator. These parameters are stored in Google Tables. Their transfer to the laboratory works html-template is carried out in encrypted form through the Google Apps script service. The simulator-parameters individualization contributes to the independence of the student's work when performing laboratory measurements in the conditions of the distance learning.

Despite the existence of multiple great resources on zk-SNARK construction, from original papers to explainers, due to the sheer number of moving parts the subject remains a black box for many. While some pieces of the puzzle are given one can not see the full picture without the missing ones. Hence the focus of this work is to shed light onto the topic with a straightforward and clean approach based on examples and answering many whys along the way so that more individuals can appreciate the state of the art technology, its innovators and ultimately the beauty of math. Paper's contribution is a simplistic exposition with a sufficient and gradually increasing level of complexity, necessary to understand zk-SNARK without any prerequisite knowledge of the subject, cryptography or advanced math. The primary goal is not only to explain how it works but why it works and how it came to be this way.

We are currently prioritizing home activities, avoiding human contact, and carrying out external activities mostly by necessity. Therefore, and due to the loss of adhesion to cultural events on the part of the population, the cultural digital transformation process has been boosted, aiming to reach interested communities through digital media. The ACOA platform supports the organization of multiple sources of information related to creative processes behind complex artworks and their trajectories over time. This information is of great interest to conservators and curators, as well as to the general public, as it allows to document changes in the artwork, from the moment it was conceived by the artist, until its most recent exhibition. This platform houses a chronological evolution of the work, through the contextual dissemination of associated multimedia content. Works by the Portuguese artist Ana Vieira (1940-2016) were chosen as case studies for the implementation of the platform.

This paper is an investigation into Cantor works about representing a function with trigonometric series, and his proofs about its uniqueness. These works are important, because they cause invention of point-set topology, and foundation of basic ideas that led Cantor to his set theory.

Problems for evaluation and impact of published scientific works and their authors are discussed. The role of citations in this process is pointed out. Different bibliometric indicators are reviewed in this connection and ways for generation of new bibliometric indices are given. The influence of different circumstances, like self-citations, number of authors, time dependence and publication types, on the evaluation and impact of scientific papers are considered. The repercussion of works citations and their content is investigated in this respect. Attention is paid also on implicit citations which are not covered by the modern bibliometrics but often are reflected in the peer reviews. Some aspects of the Web analogues of citations and new possibilities of the Internet resources in evaluating authors achievements are presented.