arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Author

Akhil

2,381 papers · 1871–2026

1967: 01968: 01969: 01970: 21971: 01972: 01973: 11974: 01975: 01976: 01977: 01978: 01979: 01980: 01981: 01982: 01983: 01984: 11985: 01986: 01987: 51988: 31989: 41990: 41991: 71992: 101993: 61994: 81995: 141996: 171997: 121998: 141999: 132000: 92001: 182002: 242003: 92004: 102005: 132006: 132007: 242008: 232009: 122010: 182011: 212012: 332013: 472014: 662015: 772016: 712017: 712018: 832019: 1052020: 1412021: 1692022: 1892023: 2122024: 2462025: 3502026: 203
1967peak 350 in 20252026

Frequent co-authors

Papers

showing 1–25 of 2,381 RSS

The Llama 3 Herd of Models

Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri +558 more

cs.AI · · cited 19,265× · arXiv ↗ · more like this ·

Modern artificial intelligence (AI) systems are powered by foundation models. This paper presents a new set of foundation models, called Llama 3. It is a herd of language models that natively support multilinguality, coding, reasoning, and tool usage. Our largest model is a dense Transformer with 405B parameters and a context window of up to 128K tokens. This paper presents an extensive empirical evaluation of Llama 3. We find that Llama 3 delivers comparable quality to leading language models such as GPT-4 on a plethora of tasks. We publicly release Llama 3, including pre-trained and post-trained versions of the 405B parameter language model and our Llama Guard 3 model for input and output safety. The paper also presents the results of experiments in which we integrate image, video, and speech capabilities into Llama 3 via a compositional approach. We observe this approach performs…

Spatializing States: Toward an Ethnography of Neoliberal Governmentality

James Ferguson, Akhil Gupta

Anthropology · American Ethnologist · · cited 2,628× · DOI ↗ · more like this ·

In this exploratory article, we ask how states come to be understood as entities with particular spatial characteristics, and how changing relations between practices of government and national territories may be challenging long‐established modes of state spatiality. In the first part of this article, we seek to identify two principles that are key to state spatialization: vertically (the state is "above"society) and encompassment (the state "encompasses" its localities). We use ethnographic evidence from a maternal health project in India to illustrate our argument that perceptions of verticality and encompassment are produced through routine bureaucratic practices. In the second part, we develop a concept of transnational governmentality as a way of grasping how new practices of government and new forms of "grassroots" politics may call into question the principles of vertical ity…

blurred boundaries: the discourse of corruption, the culture of politics, and the imagined state

Akhil Gupta

Anthropology · American Ethnologist · · cited 2,492× · DOI ↗ · more like this ·

In this article I attempt to do an ethnography of the state by examining the discourses of corruption in contemporary India. I focus on the practices of lower levels of the bureaucracy in a small north Indian town as well as on representations of the state in the mass media. Research on translocal institutions such as “the state” enables us to reflect on the limitations of participant‐observation as a technique of fieldwork. The analysis leads me to question Eurocentric distinctions between state and civil society and offers a critique of the conceptualization of “the state” as a monolithic and unitary entity. [the state, public culture, fieldwork, discourse, corruption, India]

Anthropological Locations: Boundaries and Grounds of a Field Science

Jon P. Mitchell, Akhil Gupta, James Ferguson

Political Science and International Relations · Journal of the Royal Anthropological Institute · · cited 1,562× · DOI ↗ · more like this ·

Among social sciences, anthropology relies most fundamentally on fieldwork--the long-term immersion in another way of life as basis for knowledge. In an era when anthropologists are studying topics that resist geographical localization, this book initiates a long-overdue discussion of political and epistemological implications of disciplinary commitment to fieldwork. These innovative, stimulating essays--carefully chosen to form a coherent whole--interrogate notion of the field, showing how concept is historically constructed and exploring consequences of its dominance. The essays discuss anthropological work done in places (in refugee camps, on television) or among populations (gays and lesbians, homeless people in United States) that challenge traditional boundaries of the field. The contributors suggest alternative methodologies appropriate for contemporary problems and ultimately…

Flower: A Friendly Federated Learning Research Framework

Daniel J. Beutel, Taner Topal, Akhil Mathur +8 more

cs.LG · · cited 1,436× · arXiv ↗ · more like this ·

Federated Learning (FL) has emerged as a promising technique for edge devices to collaboratively learn a shared prediction model, while keeping their training data on the device, thereby decoupling the ability to do machine learning from the need to store the data in the cloud. However, FL is difficult to implement realistically, both in terms of scale and systems heterogeneity. Although there are a number of research frameworks available to simulate FL algorithms, they do not support the study of scalable FL workloads on heterogeneous edge devices. In this paper, we present Flower -- a comprehensive FL framework that distinguishes itself from existing platforms by offering new facilities to execute large-scale FL experiments and consider richly heterogeneous FL device scenarios. Our experiments show Flower can perform FL experiments up to 15M in client size using only a pair of…

Gemini: A Family of Highly Capable Multimodal Models

Gemini Team, Rohan Anil, Sebastian Borgeaud +1348 more

cs.CL · · cited 816× · arXiv ↗ · more like this ·

This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. We discuss our approach toward post-training and deploying Gemini models…

Culture/Power/Place: Explorations in Critical Anthropology

Dorothy L. Hodgson, Akhil Gupta, James Ferguson

Sociology and Political Science · Anthropological Quarterly · · cited 562× · DOI ↗ · more like this ·

Culture/Power/Place: Explorations in Critical Anthropology. AKHIL GUPTA and JAMES FERGUSON, eds. Durham NC: Duke University Press, 1997; 361 pp. In our relentless search for theoretical guidance in understanding culture, the central conceptual cornerstone of our discipline, cultural anthropologists have long looked to other analytic realms and disciplines for insight. By counterpoising to ecology, personality, history, or now, power and we have tried to not only define the nebulous contours of what is by exploring what it is not, but sought, more importantly, to understand how is produced, reproduced, and transformed. As part of this pursuit, contemporary anthropologists have turned first to history (think, for example, of the important volume Culture/power/ history edited by Nicholas Dirks, Sherry Ortner, and Geoff Eley), and more recently to geography for assistance. The widespread…

The Song of the Nonaligned World: Transnational Identities and the Reinscription of Space in Late Capitalism

Akhil Gupta

Anthropology · Cultural Anthropology · · cited 414× · DOI ↗ · more like this ·

contingent form of organizing space in world. National identity appears to be firmly spatialized and seemingly immutable, becoming almost a natural marker of cultural and social difference. This article problematizes nationalism by juxtaposing it and other forms of spatial commitment and identity, particularly transnational ones. In so doing, it seeks to illuminate specificity of nationalism in postcolonial world. Beginning with premise that structures of feeling (R. Williams 1961:48-71) that produce a location called the nation are not identical in differently situated places, I wish to conceptualize vastly dissimilar structural positions occupied by First and Third World' nationalisms by locating them with respect to late capitalism and to postcolonial world order. Connecting such global phenomena with questions of place and identity is consonant with recent moves in anthropological…

Image restoration using a neural network

Yanjun Zhou, Rama Chellappa, Akhil Vaid +1 more

Computer Vision and Pattern Recognition · IEEE Transactions on Acoustics Speech and Signal Processing · · cited 414× · DOI ↗ · more like this ·

An approach for restoration of gray level images degraded by a known shift invariant blur function and additive noise is presented using a neural computational network. A neural network model is used to represent a possibly nonstationary image whose gray level function is the simple sum of the neuron state variables. The restoration procedure consists of two stages: estimation of the parameters of the neural network model and reconstruction of images. Owing to the model's fault-tolerant nature and computation capability, a high-quality image is obtained using this approach. A practical algorithm with reduced computational complexity is also presented. A procedure for learning the blur parameters from prototypes of original and degraded images is outlined.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Is Blockchain a Silver Bullet for Supply Chain Management? Technical Challenges and Research Opportunities

Akhil Kumar, Rong Liu, Zhe Shan

Information Systems · Decision Sciences · · cited 406× · DOI ↗ · more like this ·

ABSTRACT Blockchain technology is based on the idea of a distributed, replicated, and immutable digital ledger that enables parties to conduct business in a trustful and transparent way without the need for a central authority or intermediary. Its most popular application thus far is in payment system applications, e.g., bitcoin. This disruptive technology is expected to contribute significant business value to multiple industry sectors, including supply chain management (SCM), where it can provide greater visibility, accountability and trust in interorganizational business collaboration. In this article, we review some fundamental concepts of Hyperledger Fabric, one of the most mature permissioned blockchain implementations. Further, we use the context of a food supply chain to highlight key design and implementation challenges for blockchain, and provide a strategic assessment of its…

Understanding and Predicting Image Memorability at a Large Scale

Aditya Khosla, Akhil S. Raju, Antonio Torralba +1 more

Computer Vision and Pattern Recognition · DSpace@MIT (Massachusetts Institute of Technology) · · cited 364× · DOI ↗ · more like this ·

Progress in estimating visual memorability has been limited by the small scale and lack of variety of benchmark data. Here, we introduce a novel experimental procedure to objectively measure human memory, allowing us to build LaMem, the largest annotated image memorability dataset to date (containing 60,000 images from diverse sources). Using Convolutional Neural Networks (CNNs), we show that fine-tuned deep features outperform all other features by a large margin, reaching a rank correlation of 0.64, near human consistency (0.68). Analysis of the responses of the high-level CNN layers shows which objects and regions are positively, and negatively, correlated with memorability, allowing us to create memorability maps for each image and provide a concrete method to perform image memorability manipulation. This work demonstrates that one can now robustly estimate the memorability of…

Paradoxical Infrastructures

Cymene Howe, Jessica Lockrem, Hannah Appel +9 more

Political Science and International Relations · Science Technology & Human Values · · cited 304× · DOI ↗ · more like this ·

In recent years, a dramatic increase in the study of infrastructure has occurred in the social sciences and humanities, following upon foundational work in the physical sciences, architecture, planning, information science, and engineering. This article, authored by a multidisciplinary group of scholars, probes the generative potential of infrastructure at this historical juncture. Accounting for the conceptual and material capacities of infrastructure, the article argues for the importance of paradox in understanding infrastructure. Thematically the article is organized around three key points that speak to the study of infrastructure: ruin, retrofit, and risk. The first paradox of infrastructure, ruin, suggests that even as infrastructure is generative, it degenerates. A second paradox is found in retrofit, an apparent ontological oxymoron that attempts to bridge temporality from the…

Globalization and Postcolonial States

Akhil Gupta, Aradhana Sharma

Anthropology · Current Anthropology · · cited 297× · DOI ↗ · more like this ·

The experiences of two programs aimed at poor rural women in India suggest that postcolonial contexts might give us reason to reconsider commonly accepted characterizations of neoliberal states. An anthropological approach to the state differs from that of other disciplines by according centrality to the meanings of the everyday practices of bureaucracies and their relation to representations of the state. Such a perspective is strengthened when it integrates those meanings with political economic, social structural, and institutional approaches. Although the two programs examined here originated in different time periods (one before and the other after neoliberal reforms) and embodied very different ideologies and goals (the earlier one being a welfare program that provided tangible services and assets and the later one an empowerment program aimed at helping rural women to become…

Machine Learning to Predict Mortality and Critical Events in a Cohort of Patients With COVID-19 in New York City: Model Development and Validation

Akhil Vaid, Sulaiman Somani, Adam Russak +9 more

Artificial Intelligence · Journal of Medical Internet Research · · cited 240× · DOI ↗ · more like this ·

BACKGROUND: COVID-19 has infected millions of people worldwide and is responsible for several hundred thousand fatalities. The COVID-19 pandemic has necessitated thoughtful resource allocation and early identification of high-risk patients. However, effective methods to meet these needs are lacking. OBJECTIVE: The aims of this study were to analyze the electronic health records (EHRs) of patients who tested positive for COVID-19 and were admitted to hospitals in the Mount Sinai Health System in New York City; to develop machine learning models for making predictions about the hospital course of the patients over clinically meaningful time horizons based on patient characteristics at admission; and to assess the performance of these models at multiple hospitals and time points. METHODS: We used Extreme Gradient Boosting (XGBoost) and baseline comparator models to predict in-hospital…

SPlit: An Optimal Method for Data Splitting

V. Roshan Joseph, Akhil Vakayil

stat.ML · · cited 224× · arXiv ↗ · more like this ·

In this article we propose an optimal method referred to as SPlit for splitting a dataset into training and testing sets. SPlit is based on the method of Support Points (SP), which was initially developed for finding the optimal representative points of a continuous distribution. We adapt SP for subsampling from a dataset using a sequential nearest neighbor algorithm. We also extend SP to deal with categorical variables so that SPlit can be applied to both regression and classification problems. The implementation of SPlit on real datasets shows substantial improvement in the worst-case testing performance for several modeling methods compared to the commonly used random splitting procedure.

Squeezing Deep Learning into Mobile and Embedded Devices

Nicholas D. Lane, Sourav Bhattacharya, Akhil Mathur +3 more

Computer Vision and Pattern Recognition · IEEE Pervasive Computing · · cited 218× · DOI ↗ · more like this ·

This department provides an overview the progress the authors have made to the emerging area of embedded and mobile forms of on-device deep learning. Their work addresses two core technical questions. First, how should deep learning principles and algorithms be applied to sensor inference problems that are central to this class of computing? Second, what is required for current and future deep learning innovations to be efficiently integrated into a variety of mobile resource-constrained systems? Toward answering such questions, the authors describe phone, watch, and embedded prototypes that can locally run large-scale deep networks processing audio, images, and inertial sensor data. These prototypes are enabled with a variety of algorithmic and system-level innovations that vastly reduce conventional inference-time overhead of deep models.

Emerging security challenges in cloud computing: An insight to cloud security challenges and their mitigation

Akhil Behl

Information Systems · · cited 212× · DOI ↗ · more like this ·

Cloud computing has changed the whole picture that distributed computing used to present e.g. Grid computing, server client computing. Cloud has given a new meaning to distributed, and off-premises computing. Although, Cloud offers great benefits, it also introduces a myriad of security threats to the information and data which is now being ported from on-premises to off-premises. Where cloud computing can help organizations accomplish more by paying less (in the longer run) and breaking the physical boundaries between IT infrastructure and its users, due to openness of accessible information and data relying on trust between cloud provider and customer, heightened security threats must be overcome in order to benefit fully from this new computing exemplar. Breach in the security of any component in the cloud can be both disaster for the organization (the customer) and defacing for the…

Federated Learning of Electronic Health Records to Improve Mortality Prediction in Hospitalized Patients With COVID-19: Machine Learning Approach

Akhil Vaid, Suraj K. Jaladanki, Jie Xu +9 more

Artificial Intelligence · JMIR Medical Informatics · · cited 211× · DOI ↗ · more like this ·

Background Machine learning models require large datasets that may be siloed across different health care institutions. Machine learning studies that focus on COVID-19 have been limited to single-hospital data, which limits model generalizability. Objective We aimed to use federated learning, a machine learning technique that avoids locally aggregating raw clinical data across multiple institutions, to predict mortality in hospitalized patients with COVID-19 within 7 days. Methods Patient data were collected from the electronic health records of 5 hospitals within the Mount Sinai Health System. Logistic regression with L1 regularization/least absolute shrinkage and selection operator (LASSO) and multilayer perceptron (MLP) models were trained by using local data at each site. We developed a pooled model with combined data from all 5 sites, and a federated model that only shared…

Hierarchical quorum consensus: a new algorithm for managing replicated data

Akhil Kumar

Computer Networks and Communications · IEEE Transactions on Computers · · cited 211× · DOI ↗ · more like this ·

A novel algorithm for managing replicated data is presented. The proposed method is based on organizing the copies of an object into a logical, multilevel hierarchy, and extending the quorum consensus algorithm to such an environment. Several properties of the method are derived and optimality conditions are given for minimizing the quorum size. It is shown that, given a collection of n copies of an object, the method allows a quorum to be formed with n/sup 0.63/ copies versus ((n+1)/2) copies in the case of the majority voting algorithm. Tradeoffs between the proposed method and three other quorum-based methods are discussed, and the main features of each method are highlighted.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

COVID-19 vaccine hesitancy among medical students in India

Jyoti Jain, Suman Saurabh, Prashant Kumar +5 more

Health · Epidemiology and Infection · · cited 210× · DOI ↗ · more like this ·

The coronavirus disease 2019 (COVID-19) vaccine was launched in India on 16 January 2021, prioritising health care workers which included medical students. We aimed to assess vaccine hesitancy and factors related to it among medical students in India. An online questionnaire was filled by 1068 medical students across 22 states and union territories of India from 2 February to 7 March 2021. Vaccine hesitancy was found among 10.6%. Concern regarding vaccine safety and efficacy, lack of awareness regarding their eligibility for vaccination and lack of trust in government agencies predicted COVID-19 vaccine hesitancy among medical students. On the other hand, the presence of risk perception regarding themselves being affected with COVID-19 reduced vaccine hesitancy as well as hesitancy in participating in COVID-19 vaccine trials. Vaccine-hesitant students were more likely to derive…

Narratives of corruption

Akhil Gupta

Anthropology · Ethnography · · cited 174× · DOI ↗ · more like this ·

In this article I attempt to analyze stories about corruption for what they might reveal about the Indian state. By triangulating my own fieldwork data, a prize-winning novel written by an official of the Uttar Pradesh (U.P.) state government, and the accounts of corruption by one of the major social anthropologists of India, F.G. Bailey, I claim that narratives of corruption, and the actions of bureaucrats and agencies in relation to those narratives, are fundamental to the constitution of the state in contemporary India. I argue that the success of policies geared to providing basic needs to the poor depend on changing these narratives and thereby altering the affective relations that poor citizens have with the state.

Vision-Language Models as Success Detectors

Yuqing Du, Ksenia Konyushkova, Misha Denil +5 more

cs.CV · · cited 173× · arXiv ↗ · more like this ·

Detecting successful behaviour is crucial for training intelligent agents. As such, generalisable reward models are a prerequisite for agents that can learn to generalise their behaviour. In this work we focus on developing robust success detectors that leverage large, pretrained vision-language models (Flamingo, Alayrac et al. (2022)) and human reward annotations. Concretely, we treat success detection as a visual question answering (VQA) problem, denoted SuccessVQA. We study success detection across three vastly different domains: (i) interactive language-conditioned agents in a simulated household, (ii) real world robotic manipulation, and (iii) "in-the-wild" human egocentric videos. We investigate the generalisation properties of a Flamingo-based success detection model across unseen language and visual changes in the first two domains, and find that the proposed method is able to…

A foundation model for clinician-centered drug repurposing

Kexin Huang, Payal Chandak, Qianwen Wang +7 more

Computational Theory and Mathematics · medRxiv · · cited 171× · DOI ↗ · more like this ·

Drug repurposing - identifying new therapeutic uses for approved drugs - is often serendipitous and opportunistic, expanding the use of drugs for new diseases. The clinical utility of drug repurposing AI models remains limited because the models focus narrowly on diseases for which some drugs already exist. Here, we introduce TXGNN, a graph foundation model for zero-shot drug repurposing, identifying therapeutic candidates even for diseases with limited treatment options or no existing drugs. Trained on a medical knowledge graph, TXGNN utilizes a graph neural network and metric-learning module to rank drugs as potential indications and contraindications across 17,080 diseases. When benchmarked against eight methods, TXGNN improves prediction accuracy for indications by 49.2% and contraindications by 35.1% under stringent zero-shot evaluation. To facilitate model interpretation, TXGNN's…

MCM-GPU

Akhil Arunkumar, Evgeny Bolotin, Benjamin Cho +6 more

Hardware and Architecture · · cited 160× · DOI ↗ · more like this ·

Historically, improvements in GPU-based high performance computing have been tightly coupled to transistor scaling. As Moore's law slows down, and the number of transistors per die no longer grows at historical rates, the performance curve of single monolithic GPUs will ultimately plateau. However, the need for higher performing GPUs continues to exist in many domains. To address this need, in this paper we demonstrate that package-level integration of multiple GPU modules to build larger logical GPUs can enable continuous performance scaling beyond Moore's law. Specifically, we propose partitioning GPUs into easily manufacturable basic GPU Modules (GPMs), and integrating them on package using high bandwidth and power efficient signaling technologies. We lay out the details and evaluate the feasibility of a basic Multi-Chip-Module GPU (MCM-GPU) design. We then propose three…

W-RBAC — A Workflow Security Model Incorporating Controlled Overriding of Constraints

Jacques Wainer, Paulo Barthelmess, Akhil Kumar

Sociology and Political Science · International Journal of Cooperative Information Systems · · cited 158× · DOI ↗ · more like this ·

This paper presents a pair of role-based access control models for workflow systems, collectively known as the W-RBAC models. The first of these models, W0-RBAC is based on a framework that couples a powerful RBAC-based permission service and a workflow component with clear separation of concerns for ease of administration of authorizations. The permission service is the focus of the work, providing an expressive logic-based language for the selection of users authorized to perform workflow tasks, with preference ranking. W1-RBAC extends the basic model by incorporating exception handling capabilities through controlled and systematic overriding of constraints.