The paper considers Percival Everett’s novel James (2024) and its peculiar technique as an example of resignification and a contemporary reworking of Twain’s Adventures of Huckleberry Finn and American minstrel shows. The theoretical framework is based on Judith Butler’s idea of resignification and Eric Lott’s considerations of the notion of minstrelsy, highlighting the connections between the two. As the key aspect of both Everett’s novel and a minstrel show is performativity/performance, this notion is discussed in detail with a special emphasis on its use in the novel. We argue that Everett’s innovative and subversive use of language, and his decision to tell the story of the birth of a hero from the point of view of Jim/James, the (in)famous stereotypical slave character of Twain’s narrative, represents an instance of anti-minstrel fiction and thus sets new standards in contemporary (African) American literature. We argue that Jim’s/James’ journey from a runaway slave and a comic relief in Twain’s classic to a hero of his own narrative can be interpreted as an example of resignification, i.e., ‘enacted critique’, a deliberate, deviant, mis-performance (Loxley, 2006, 127) which paves the way for new roles and meanings in contemporary (African) American literature.
Viviana Paola Delgado Sánchez, Ana Zorio-Grima, Paloma Merello
This study investigates the factors influencing the adoption of assurance practices in sustainability reporting among leading companies across 42 countries from 2019 to 2022. Using panel data models, it examines the assurance lag, duality between the choice of audit firm and assuror for financial and environmental, social, and governance (ESG) assurance, and level of assurance (reasonable vs. limited). The results indicate that auditor–assuror duality may reduce the assurance lag through improved consistency and efficiency. However, this choice is not driven by expected benefits such as the inclusion of ESG information in annual reports or a preference for audit firms over consultants as assurors. Additionally, when audit firms follow specific assurance standards, there is evidence of a negative impact on the percentage of ESG information verified at the reasonable assurance level. This apparent negative impact is probably due to conservative approaches and strict methodological requirements. The findings offer insight to support decision-making for companies and regulators in enhancing transparency and trust in sustainability reports. This insight is particularly relevant in light of changes in European Union (EU) regulations that may impact assurance trends. Specifically, the new Corporate Sustainability Reporting Directive allows for different options that may be transposed differently by EU member states. The study thus has valuable implications regarding the future regulatory environment in many contexts.
History of scholarship and learning. The humanities, Social sciences (General)
Pemahaman terhadap tipe kepribadian menjadi mutlak pada kondisi digitalisasi dan hybrid working. Tipe kepribadian yang umum dikenal saat ini adalah introver dan ekstrover. Organisasi yang tidak mampu memahami tipe kepribadian karyawan, akan berdampak pada penurunan motivasi dan kinerja karyawan. Salah satu cara mengklasifikasikan tipe kepribadian pegawai adalah dengan pendekatan machine learning. Evaluasi terhadap beberapa hasil pendekatan machine learning, akan memberikan model dengan kinerja terbaik yang mampu mengklasifikasikan tipe kepribadian. Model Naïve Bayes menjadi model terbaik pada klasfikasi tipe kepribadian ini dengan nilai accuracy sebesar 93,41%, lebih tinggi dibandingkan model lainnya. Penelitian ini diharapkan menambah wawasan ilmu pengetahuan pada human resources analitik dan memberikan informasi klasifikasi tipe kepribadian karyawan bagi organisasi.
Weiqin Chen, Xinjie Zhang, Dharmashankar Subramanian
et al.
Transformer models (TMs) have exhibited remarkable in-context reinforcement learning (ICRL) capabilities, allowing them to generalize to and improve in previously unseen environments without re-training or fine-tuning. This is typically accomplished by imitating the complete learning histories of a source RL algorithm over a substantial amount of pretraining environments, which, however, may transfer suboptimal behaviors inherited from the source algorithm/dataset. Therefore, in this work, we address the issue of inheriting suboptimality from the perspective of dataset preprocessing. Motivated by the success of the weighted empirical risk minimization, we propose a simple yet effective approach, learning history filtering (LHF), to enhance ICRL by reweighting and filtering the learning histories based on their improvement and stability characteristics. To the best of our knowledge, LHF is the first approach to avoid source suboptimality by dataset preprocessing, and can be combined with the current state-of-the-art (SOTA) ICRL algorithms. We substantiate the effectiveness of LHF through a series of experiments conducted on the well-known ICRL benchmarks, encompassing both discrete environments and continuous robotic manipulation tasks, with three SOTA ICRL algorithms (AD, DPT, DICP) as the backbones. LHF exhibits robust performance across a variety of suboptimal scenarios, as well as under varying hyperparameters and sampling strategies. Notably, the superior performance of LHF becomes more pronounced in the presence of noisy data, indicating the significance of filtering learning histories.
Curriculum reform has revolved from production-based to knowledge-based. This trajectory compelled reformists to be deliberate in generating knowledge toward globalized curriculum practices that value humanity for good. This empirical paper reflected on the lived experiences of postgraduate students in a higher learning institution in their becoming. Postgraduate students are expected to complete their studies by working in a space buffeted by silo-thinking, academic jealousy, and volatile relations. This study attempted to answer the following question. What curriculum practices and innovations can make the world a better place for all? Africana Critical Theory was used to make sense of the lived experiences of a postgraduate student. Eight postgraduate students registered for masters and doctoral studies, and five supervisory teams, operating through multiple artificial intelligence in the form of the Yammer tool ranging from smartphones, laptops were engaged numerous times to enable sharing, expressing, and showing casing their new emergent identity in a virtual participatory action research approach, online discussion. Webinars were the primary source for generating data. The generated data was recorded and automatically generated into text through Fireflies’ software. Critical Discourse Analysis was used to arrive at the following findings: Curriculum practices unraveled the hidden curriculum that humans in our current state, that cybernetics exists around us, and in simpler forms than futuristic visions. Cyberspace has created profound variations in human consciousness and social identity. These findings imply that second-life realities are beneficial in a postgraduate context. Keywords: Curriculum Practices, Yammers Tool, Consciousness, Postgraduate
Introduction. The article analyzes a little–studied problem in historiography – khotons (nomadic settlements) the Kalmyk Khanate of the XVIII century, which at the same time were the smallest administrative units.Materials and methods. The study is based on an extensive set of archival materials, primarily clerical correspondence, which somehow deposited information about the khotons of the Kalmyk Khanate of the XVIII century, the number of people in them, the degree of kinship, etc. Historical-genetic, comparative-historical, functional and descriptive methods were used in the analysis of source and bibliographic material and in writing the text of the article.Analysis. The analysis showed that a significant part of khotons consisted of 10-15 kibits (families), which were connected with each other by kinship relations, including those who were at different levels of kinship relations. However, archival documents show that in the 18th century, in addition to ordinary khotons consisting of related families, nomadic settlements began to appear in the Kalmyk Khanate, consisting of families or individuals not related by kinship and even belonging to different sub-ethnic groups. Sometimes, for various reasons, several khotons could unite and form a separate group of several dozens of kibits, and in some cases, due to extraordinary circumstances, they could form groups even of several hundred kibits, though for a short period. In addition to the official authorities, khotons and other nomadic groups of khoton type had their own internal public self-government built on the life experience and authority of their managers.Results. Thus, it can be concluded that in the socially stratified Kalmyk society of the XVIII century tribal relations began to deform under the influence of the prevailing circumstances.
Law, History of scholarship and learning. The humanities
Winda Widyanty, Dian Primanita Oktasari, Sugeng Santoso
et al.
The quality of life (QoL) of workers during the Covid-19 pandemic is an important issue that must be considered. Unfortunately, research related to the QoL of workers during the Covid-19 pandemic for the non-health sector is still very limited. Moreover, no one has comprehensively investigated QoL involving not only the perceived threat of Covid-19, Covid-19-related workplace policy, and job insecurity but also digital literacy, perceived organizational support (POS) during Covid-19, quality culture, and safety culture. Therefore, to fill the gap in the literature, this study studied QoL by involving perceived threat of Covid-19, Covid-19 related workplace policy, job insecurity, digital literacy, POS, quality culture, and safety culture. Quantitative research method was carried out in this research. Data collection was conducted through an online survey. The research respondents were 181 non-health sector workers in Indonesia. SEM-PLS was used as an analytical tool. The results showed that QoL was directly and positively affected by POS and safety culture. In addition, QoL was also indirectly affected by Covid-19-related workplace policy, quality culture and safety culture by post. However, several factors, namely the perceived threat of Covid-19, job insecurity, and digital literacy did not have a significant effect on the QoL of non-health sector workers during the Covid-19 pandemic. In addition, this research also found that quality culture did not affect Covid-19 related workplace policy and job insecurity. The perceived threat of Covid-19 was not affected by the Covid-19-related workplace policy and safety culture. Job insecurity was affected by the perceived threat of Covid-19.
History of scholarship and learning. The humanities, Social Sciences
Theodore Papamarkou, Tolga Birdal, Michael Bronstein
et al.
Topological deep learning (TDL) is a rapidly evolving field that uses topological features to understand and design deep learning models. This paper posits that TDL is the new frontier for relational learning. TDL may complement graph representation learning and geometric deep learning by incorporating topological concepts, and can thus provide a natural choice for various machine learning settings. To this end, this paper discusses open problems in TDL, ranging from practical benefits to theoretical foundations. For each problem, it outlines potential solutions and future research opportunities. At the same time, this paper serves as an invitation to the scientific community to actively participate in TDL research to unlock the potential of this emerging field.
Personalized federated learning (PFL) offers a flexible framework for aggregating information across distributed clients with heterogeneous data. This work considers a personalized federated learning setting that simultaneously learns global and local models. While purely local training has no communication cost, collaborative learning among the clients can leverage shared knowledge to improve statistical accuracy, presenting an accuracy-communication trade-off in personalized federated learning. However, the theoretical analysis of how personalization quantitatively influences sample and algorithmic efficiency and their inherent trade-off is largely unexplored. This paper makes a contribution towards filling this gap, by providing a quantitative characterization of the personalization degree on the tradeoff. The results further offers theoretical insights for choosing the personalization degree. As a side contribution, we establish the minimax optimality in terms of statistical accuracy for a widely studied PFL formulation. The theoretical result is validated on both synthetic and real-world datasets and its generalizability is verified in a non-convex setting.
Ellen Novoseller, Vinicius G. Goecks, David Watkins
et al.
In machine learning for sequential decision-making, an algorithmic agent learns to interact with an environment while receiving feedback in the form of a reward signal. However, in many unstructured real-world settings, such a reward signal is unknown and humans cannot reliably craft a reward signal that correctly captures desired behavior. To solve tasks in such unstructured and open-ended environments, we present Demonstration-Inferred Preference Reinforcement Learning (DIP-RL), an algorithm that leverages human demonstrations in three distinct ways, including training an autoencoder, seeding reinforcement learning (RL) training batches with demonstration data, and inferring preferences over behaviors to learn a reward function to guide RL. We evaluate DIP-RL in a tree-chopping task in Minecraft. Results suggest that the method can guide an RL agent to learn a reward function that reflects human preferences and that DIP-RL performs competitively relative to baselines. DIP-RL is inspired by our previous work on combining demonstrations and pairwise preferences in Minecraft, which was awarded a research prize at the 2022 NeurIPS MineRL BASALT competition, Learning from Human Feedback in Minecraft. Example trajectory rollouts of DIP-RL and baselines are located at https://sites.google.com/view/dip-rl.
Konstantin Burlachenko, Abdulmajeed Alrowithi, Fahad Ali Albalawi
et al.
Traditional AI methodologies necessitate centralized data collection, which becomes impractical when facing problems with network communication, data privacy, or storage capacity. Federated Learning (FL) offers a paradigm that empowers distributed AI model training without collecting raw data. There are different choices for providing privacy during FL training. One of the popular methodologies is employing Homomorphic Encryption (HE) - a breakthrough in privacy-preserving computation from Cryptography. However, these methods have a price in the form of extra computation and memory footprint. To resolve these issues, we propose an innovative framework that synergizes permutation-based compressors with Classical Cryptography, even though employing Classical Cryptography was assumed to be impossible in the past in the context of FL. Our framework offers a way to replace HE with cheaper Classical Cryptography primitives which provides security for the training process. It fosters asynchronous communication and provides flexible deployment options in various communication topologies.
Christian Bök é um escritor e poeta conceitual canadense, professor de Escrita Criativa na Universidade de Melbourne, na Austrália. Na esteira de outros poetas conceituais que deram início ao campo de produção da arte transgênica, Bök engendrou um projeto híbrido entre as áreas da poesia e da biologia a fim de criar um poema vivo e imortal. Através de uma formação autodidata nas áreas de biologia molecular, biogenética e engenharia genética, o poeta desenvolveu, em teoria, uma forma de escrever um pequeno poema e introduzi-lo no DNA de uma bactéria resistente ao fogo, à água, à radiação e à própria explosão do Sol, capaz de sobreviver, inclusive, no espaço sideral. Além disso, parte do seu projeto era induzir a bactéria a ler o poema original e a produzir um verso poético em resposta. Este artigo descreve como Christian Bök comprovou em laboratório a viabilidade da sua teoria ao escrever um verso no DNA de uma bactéria modelo para estudos genéticos, a Escherichia coli, e obter dela, na sequência, um verso em resposta.
History of scholarship and learning. The humanities, Philology. Linguistics
Different models can provide differing levels of fidelity when a robot is planning. Analytical models are often fast to evaluate but only work in limited ranges of conditions. Meanwhile, physics simulators are effective at modeling complex interactions between objects but are typically more computationally expensive. Learning when to switch between the various models can greatly improve the speed of planning and task success reliability. In this work, we learn model deviation estimators (MDEs) to predict the error between real-world states and the states outputted by transition models. MDEs can be used to define a model precondition that describes which transitions are accurately modeled. We then propose a planner that uses the learned model preconditions to switch between various models in order to use models in conditions where they are accurate, prioritizing faster models when possible. We evaluate our method on two real-world tasks: placing a rod into a box and placing a rod into a closed drawer.
Inductive transfer learning aims to learn from a small amount of training data for the target task by utilizing a pre-trained model from the source task. Most strategies that involve large-scale deep learning models adopt initialization with the pre-trained model and fine-tuning for the target task. However, when using over-parameterized models, we can often prune the model without sacrificing the accuracy of the source task. This motivates us to adopt model pruning for transfer learning with deep learning models. In this paper, we propose PAC-Net, a simple yet effective approach for transfer learning based on pruning. PAC-Net consists of three steps: Prune, Allocate, and Calibrate (PAC). The main idea behind these steps is to identify essential weights for the source task, fine-tune on the source task by updating the essential weights, and then calibrate on the target task by updating the remaining redundant weights. Under the various and extensive set of inductive transfer learning experiments, we show that our method achieves state-of-the-art performance by a large margin.
Simon Guiroy, Christopher Pal, Gonçalo Mordido
et al.
Meta-Learning algorithms for few-shot learning aim to train neural networks capable of generalizing to novel tasks using only a few examples. Early-stopping is critical for performance, halting model training when it reaches optimal generalization to the new task distribution. Early-stopping mechanisms in Meta-Learning typically rely on measuring the model performance on labeled examples from a meta-validation set drawn from the training (source) dataset. This is problematic in few-shot transfer learning settings, where the meta-test set comes from a different target dataset (OOD) and can potentially have a large distributional shift with the meta-validation set. In this work, we propose Activation Based Early-stopping (ABE), an alternative to using validation-based early-stopping for meta-learning. Specifically, we analyze the evolution, during meta-training, of the neural activations at each hidden layer, on a small set of unlabelled support examples from a single task of the target tasks distribution, as this constitutes a minimal and justifiably accessible information from the target problem. Our experiments show that simple, label agnostic statistics on the activations offer an effective way to estimate how the target generalization evolves over time. At each hidden layer, we characterize the activation distributions, from their first and second order moments, then further summarized along the feature dimensions, resulting in a compact yet intuitive characterization in a four-dimensional space. Detecting when, throughout training time, and at which layer, the target activation trajectory diverges from the activation trajectory of the source data, allows us to perform early-stopping and improve generalization in a large array of few-shot transfer learning settings, across different algorithms, source and target datasets.
This study is concerned with investigating the deliberative goals behind the use of popular slang expressions by the preachers of the Husseini pulpit in the context of delivering lectures addressed to the Iraqi people and the Islamic public in general. These preachers are considered social reformers, in addition to being descended from the religious establishment, as they put forward religious ideas, most of which derive from the Husseini issue, and use them as a basis for calling for reform. Hence, the popular colloquial language is employed by these Husseini preachers in their lectures and sermons, which are characterized - as the case may be - by a very formal context. However, after presenting a theoretical aspect of the deliberative aspect of the popular vernacular language, and after examining the deliberative functions in a number of lectures and by different preachers. The most commonly used and frequently used colloquial expressions were selected and then placed in a questionnaire according to the approved evidence chosen for this st
History of scholarship and learning. The humanities, Arts in general
Learning from raw data input, thus limiting the need for manual feature engineering, is one of the key components of many successful applications of machine learning methods. While machine learning problems are often formulated on data that naturally translate into a vector representation suitable for classifiers, there are data sources, for example in cybersecurity, that are naturally represented in diverse files with a unifying hierarchical structure, such as XML, JSON, and Protocol Buffers. Converting this data to vector (tensor) representation is generally done by manual feature engineering, which is laborious, lossy, and prone to human bias about the importance of particular features. Mill and JsonGrinder is a tandem of libraries, which fully automates the conversion. Starting with an arbitrary set of JSON samples, they create a differentiable machine learning model capable of infer from further JSON samples in their raw form.
Abstract Using citation analysis, we consider the role of gender in citation practices in conference special issues of Digital Scholarship in the Humanities. Our examination of citations in Digital Humanities conference special issues from 2006 to 2015 demonstrates gender bias in citational practices. This bias is consistent with broader trends in citational politics across the academy more broadly but is a threat to equity and justice within the scholarly community. We further offer proposals for improving citational practices to resist gender bias. Quantifying the impact of gender on citations, we argue, is one approach to understanding gender inequalities within digital humanities communities and to generating solutions to promote the broadest representation of digital humanities scholarship in scholarly communications.