arXiv.itsnick.co

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

Author

Pushkar

723 papers · 1753–2026

1967: 01968: 01969: 01970: 11971: 01972: 01973: 01974: 01975: 01976: 01977: 01978: 11979: 01980: 01981: 01982: 01983: 01984: 01985: 01986: 01987: 01988: 11989: 01990: 01991: 01992: 01993: 01994: 01995: 11996: 11997: 11998: 11999: 22000: 52001: 52002: 42003: 82004: 52005: 82006: 102007: 142008: 162009: 172010: 182011: 172012: 202013: 182014: 132015: 302016: 262017: 262018: 192019: 372020: 422021: 452022: 462023: 612024: 672025: 922026: 42
1967peak 92 in 20252026

Frequent co-authors

Papers

showing 1–25 of 723 RSS

Llama 2: Open Foundation and Fine-Tuned Chat Models

Hugo Touvron, Louis Martin, Kevin Stone +65 more

cs.CL · · cited 18,418× · arXiv ↗ · more like this ·

In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama 2-Chat, are optimized for dialogue use cases. Our models outperform open-source chat models on most benchmarks we tested, and based on our human evaluations for helpfulness and safety, may be a suitable substitute for closed-source models. We provide a detailed description of our approach to fine-tuning and safety improvements of Llama 2-Chat in order to enable the community to build on our work and contribute to the responsible development of LLMs.

Harmonic coordinates for character articulation

Pushkar Joshi, Mark Meyer, Tony DeRose +2 more

Computer Graphics and Computer-Aided Design · ACM Transactions on Graphics · · cited 394× · DOI ↗ · more like this ·

In this paper we consider the problem of creating and controlling volume deformations used to articulate characters for use in high-end applications such as computer generated feature films. We introduce a method we call harmonic coordinates that significantly improves upon existing volume deformation techniques. Our deformations are controlled using a topologically flexible structure, called a cage, that consists of a closed three dimensional mesh. The cage can optionally be augmented with additional interior vertices, edges, and faces to more precisely control the interior behavior of the deformation. We show that harmonic coordinates are generalized barycentric coordinates that can be extended to any dimension. Moreover, they are the first system of generalized barycentric coordinates that are non-negative even in strongly concave situations, and their magnitude falls off with…

Motivation, Interest, and Positive Affect in Traditional and Nontraditional Undergraduate Students

Dorothea Bye, Dolores Pushkar, Michael Conway

Education · Adult Education Quarterly · · cited 387× · DOI ↗ · more like this ·

This study compares affective and motivational components of academic life for traditional and nontraditional university undergraduates. Traditional students are defined as those aged 21 and younger, who are most likely to have followed an unbroken linear path through the education system, whereas nontraditional students are defined as those aged 28 and older, for whom the undergraduate experience is not necessarily age normative. A total of 300 undergraduates ranging in age from 18 to 60 years were assessed on measures of intrinsic and extrinsic motivation to learn, interest, and positive affect. Nontraditional students reported higher levels of intrinsic motivation for learning than did traditional students. Intrinsic motivation correlated with positive affect more strongly for nontraditional than for traditional students. For all students, interest and age emerged as significant…

EndRE: An End-System Redundancy Elimination Service for Enterprises

Bhavish Aggarwal, Aditya Akella, Ashok An +5 more

Computer Networks and Communications · · cited 210× · OpenAlex ↗ · more like this ·

In many enterprises today, middleboxes called WAN optimizers are being deployed across WAN access links in order to eliminate redundancy in network traffic and reduce WAN access costs. In this paper, we present the design and implementation of EndRE, an alternate approach where redundancy elimination is provided as an end system service. Unlike middleboxes, such an approach benefits both end-to-end encrypted traffic as well as traffic on last-hop wireless links to mobile devices. EndRE needs to be fast, adaptive and parsimonious in memory usage in order to opportunistically leverage resources on end hosts. Thus, we design a new fingerprinting scheme called SampleByte that is much faster than Rabin fingerprinting while delivering similar compression gains. Unlike Rabin, SampleByte can also adapt its CPU usage depending on server load. Further, we introduce optimizations to reduce server…

Online bipartite matching with unknown distributions

Chinmay Karande, Aranyak Mehta, Pushkar Tripathi

Computer Networks and Communications · · cited 203× · DOI ↗ · more like this ·

We consider the online bipartite matching problem in the unknown distribution input model. We show that the Ranking algorithm of [KVV90] achieves a competitive ratio of at least 0.653. This is the first analysis to show an algorithm which breaks the natural 1 - 1/e -barrier' in the unknown distribution model (our analysis in fact works in the stricter, random order model) and answers an open question in [GM08]. We also describe a family of graphs on which Ranking does no better than 0.727 in the random order model. Finally, we show that for graphs which have k > 1 disjoint perfect matchings, Ranking achieves a competitive ratio of at least 1 - √(1/k - 1/k2 + 1/n) -- in particular Ranking achieves a factor of 1 - o(1) for graphs with ω(1) disjoint perfect matchings.

Homomorphic Encryption

Monique Ogburn, Claude Turner, Pushkar Dahal

Artificial Intelligence · Procedia Computer Science · · cited 142× · DOI ↗ · more like this ·

The study of homomorphic encryption techniques has led to significant advancements in the computing domain, particularly in the sphere of cloud computing. Homomorphic encryption provides a means for securely transmitting and storing confidential information across and in a computer system. The aim of this paper is to discuss the concepts and significance of homomorphic encryption along with the subdivisions and limitations associated with this type of encryption scheme. Recent studies conducted on the topic of homomorphic encryption are highlighted and some customary models of homomorphism are demonstrated. We also developed a proof of concept algorithm that demonstrates a practical use for a homomorphic encryption technique, the results of our algorithm are provided. The applications of homomorphic encryption methods are vast outside of the computational realm, and its purpose in other…

The Joint Estimation of Child Participation in Schooling and Employment: Comparative Evidence from Three Continents

Pushkar Maitra, Ranjan Ray

Safety Research · Oxford Development Studies · · cited 134× · DOI ↗ · more like this ·

This paper uses data from Peru, Pakistan and Ghana to simultaneously analyse child labour and child schooling, and compares them between these countries .We use a multinomial logit estimation procedure that analyses the participation and non participation of children in schooling and in employment and, in particular, allows the possibility that a child combines schooling with employment or does neither. We also use an ordered probit estimation procedure based on a ranking of the various child schooling/employment/non schooling/non employment outcomes. The results point to both similarities and striking dissimilarities in the nature of child labour and child schooling between the chosen countries. For example, in Pakistan, but not in Peru, the girl child’s ordering of schooling/employment outcomes shows her at a position of extreme disadvantage. Household poverty discourages a child from…

Social Learning and Norms in a Public Goods Experiment with Inter-Generational Advice1

Ananish Chaudhuri, Sara Graziano, Pushkar Maitra

Safety Research · The Review of Economic Studies · · cited 132× · DOI ↗ · more like this ·

We study a linear public goods game using an inter-generational approach. Subjects in one generation leave advice for the succeeding generation via free-form messages. Such advice can be private knowledge (advice left by one player in generation t is given only to his or her immediate successor in generation t + 1), public knowledge (advice left by players of generation t is made available to all members of generation t + 1), and common knowledge (where the advice is not only public but is also read aloud by the experimenter). Common knowledge of advice generates a process of social learning that leads to high contributions and less free-riding. This behaviour is sustained by advice that is generally exhortative, suggesting high contributions, which in turn creates optimistic beliefs among subjects about others' contributions. We suggest that socially connected communities may achieve…

Distance learning in higher education during pandemic: challenges and opportunities

Pushkar Dubey, Deepak Pandey

Computer Science Applications · International Journal of Indian Psychology · · cited 100× · DOI ↗ · more like this ·

Sudden Breakout of Covid-19 Pandemic has called for a pause in the ways and means of learning in higher educational institutions. The situation has prevented the learning actions for the students. The activities of the higher educational institutions are being affected and it is unknown how far the situation will last.  At the prevailing circumstances it is important to shift to distance platform of learning through the online process.  In a country like India with diverse learner group and infrastructural challenges, the possibilities of learning through online mode has always been under debate. The present paper highlights the common problems faced by the higher educational institutions and the possible measures which can be initiated in the present situation for uninterrupted learning to occur.

Tackling Online Abuse: A Survey of Automated Abuse Detection Methods

Pushkar Mishra, Helen Yannakoudakis, Ekaterina Shutova

cs.CL · · cited 93× · arXiv ↗ · more like this ·

Abuse on the Internet represents an important societal problem of our time. Millions of Internet users face harassment, racism, personal attacks, and other types of abuse on online platforms. The psychological effects of such abuse on individuals can be profound and lasting. Consequently, over the past few years, there has been a substantial research effort towards automated abuse detection in the field of natural language processing (NLP). In this paper, we present a comprehensive survey of the methods that have been proposed to date, thus providing a platform for further development of this area. We describe the existing datasets and review the computational approaches to abuse detection, analyzing their strengths and limitations. We discuss the main trends that emerge, highlight the challenges that remain, outline possible solutions, and propose guidelines for ethics and…

Schooling and Educational Attainment: Evidence from Bangladesh

Pushkar Maitra

Safety Research · Education Economics · · cited 91× · DOI ↗ · more like this ·

Education and human capital accumulation are essential components of economic development. This paper attempts to identify some of the individual and household level characteristics that affect the demand for schooling in Bangladesh. We examine (1) current enrolment status of children aged 6 -12 and (2) the highest grade attained for children aged 13 -24. The rst is estimated using a standard probit model and the second using a censored ordered probit model. Estimation results show that there is no gender differential in current enrolment status but grade attainment is higher by girls, relative to boys. Increases in the permanent income of the household is always associated with an increase in educational attainment. Parental education generally has a positive and statistically signicant effect on the educational attainment of children and mother's education has a stronger effect on…

K-nearest correlated neighbor classification for Indian sign language gesture recognition using feature fusion

Bhumika Gupta, Pushkar Shukla, Ankush Mittal

Human-Computer Interaction · · cited 86× · DOI ↗ · more like this ·

A sign language recognition system is an attempt to bring the speech and the hearing impaired community closer to more regular and convenient forms of communication. Thus, this system requires to recognize the gestures from a sign language and convert them to a form easily understood by the hearing. The model that has been proposed in this paper recognizes static images of the signed alphabets in the Indian Sign Language. Unlike the alphabets in other sign languages like the American Sign Language and the Chinese Sign language, the ISL alphabet are both single-handed and double-handed. Hence, to make recognition easier the model first categorizes them as single-handed or double-handed. For both categories two kinds of features, namely HOG and SIFT, are extracted for a set of training images and are combined in a single matrix. After which, HOG and SIFT features for the input test image…

Abusive Language Detection with Graph Convolutional Networks

Pushkar Mishra, Marco Del Tredici, Helen Yannakoudakis +1 more

cs.CL · · cited 85× · arXiv ↗ · more like this ·

Abuse on the Internet represents a significant societal problem of our time. Previous research on automated abusive language detection in Twitter has shown that community-based profiling of users is a promising technique for this task. However, existing approaches only capture shallow properties of online communities by modeling follower-following relationships. In contrast, working with graph convolutional networks (GCNs), we present the first approach that captures not only the structure of online communities but also the linguistic behavior of the users within them. We show that such a heterogeneous graph-structured modeling of communities significantly advances the current state of the art in abusive language detection.

Quantum K-theory of Quiver Varieties and Many-Body Systems

Peter Koroteev, Petr P. Pushkar, Andrey Smirnov +1 more

math.AG · · cited 80× · arXiv ↗ · more like this ·

We define quantum equivariant K-theory of Nakajima quiver varieties. We discuss type A in detail as well as its connections with quantum XXZ spin chains and trigonometric Ruijsenaars-Schneider models. Finally we study a limit which produces a K-theoretic version of results of Givental and Kim, connecting quantum geometry of flag varieties and Toda lattice.

Approximability of combinatorial problems with multi-agent submodular cost functions

Gagan Goel, Chinmay Karande, Pushkar Tripathi +1 more

Computational Theory and Mathematics · ACM SIGecom Exchanges · · cited 78× · DOI ↗ · more like this ·

Applications in complex systems such as the Internet have spawned a recent interest in studying situations involving multiple agents with their individual cost or utility functions. In this paper, we introduce an algorithmic framework for studying combinatorial optimization problems in the presence of multiple agents with submodular cost functions. We study several fundamental covering problems in this framework and establish upper and lower bounds on their approximability.

Household Expenditure Patterns and Gender Bias: Evidence from Selected Indian States

Geoffrey Lancaster, Pushkar Maitra, Ranjan Ray

Gender Studies · Oxford Development Studies · · cited 74× · DOI ↗ · more like this ·

This paper uses Indian data to investigate the existence and nature of gender bias in the intra-household allocation of expenditure. An extended version of the collective household model is estimated where the welfare weights, i.e. the bargaining power of the adult decision-makers, are simultaneously determined with the household's expenditure outcomes. Significant gender bias is detected in some items, most notably in education, and it is found that the bias is considerably stronger in the more economically backward regions of the country. It is also found that the results of the test of gender bias vary sharply between households at different levels of adult literacy. This is particularly true of household spending on education. The gender bias in the case of this item is, generally, more likely to prevail in households with low levels of adult educational attainment than in more…

FACES: Friend-Based Ad Hoc Routing Using Challenges to Establish Security in MANETs Systems

Sanjay Kumar Dhurandher, Mohammad S. Obaidat, Karan Verma +2 more

Computer Networks and Communications · IEEE Systems Journal · · cited 70× · DOI ↗ · more like this ·

Friend based Ad hoc routing using Challenges to Establish Security (FACES) is an algorithm to provide secure routing in ad hoc mobile networks. We propose this scheme that has been drawn from a network of friends in real life scenarios. The algorithm works by sending challenges and sharing friend Lists to provide a list of trusted nodes to the source node through which data transmission finally takes place. The nodes in the friend list are rated on the basis of the amount of data transmission they accomplish and their friendship with other nodes in the network. The account of friendship of a node with other nodes in the network is obtained through the Share Your Friends process which is a periodic event in the network. As a result of this scheme of operation, the network is able to effectively isolate the malicious nodes which are left with no role to play in the ad hoc network. One…

Test-Driving the Future: How Design Competitions Are Changing Innovation

Joseph Lampel, Pushkar Jha, Ajay Bhalla

Computer Science Applications · Academy of Management Perspectives · · cited 68× · DOI ↗ · more like this ·

Executive Overview As organizations realize the potential of “open innovation” models, design competitions—target-setting events that offer monetary awards and other benefits to contestants—are regaining popularity as an innovation tool. In this paper we look at the innovation agendas of organizations and individuals who sponsor and organize design competitions. We then examine the architecture and governance of such competitions, and explore how open innovation and crowdsourcing, in combination with online platforms, have transformed design competitions. Finally, we look at the evolution of design competitions and highlight their expanding scope and complexity.

Effect of off-target verbosity on communication efficiency in a referential communication task.

Tannis Y. Arbuckle, Michiko Nohara-LeClair, Dolores Pushkar

Artificial Intelligence · Psychology and Aging · · cited 66× · DOI ↗ · more like this ·

The referential communication task was used to see if high off-target verbosity (OTV), defined as excessive speech that is lacking in focus, negatively affects communication of nonautobiographical information. The task required 1 individual (the director) to communicate descriptions of abstract figures to another (the matcher). Out of 455 adults aged 63 to 93 who were screened for OTV, 27 directors were drawn from each of the top and bottom 15% of the range of OTV scores and 26 directors and all 80 matchers from the middle 50%. High OTV directors were less efficient communicators about the figures and showed poorer inhibitory control but did not intrude personal information in their speech. The implications of the findings for the inhibitory deficit and pragmatic change explanations of OTV are discussed.

Neural Character-based Composition Models for Abuse Detection

Pushkar Mishra, Helen Yannakoudakis, Ekaterina Shutova

cs.CL · · cited 65× · arXiv ↗ · more like this ·

The advent of social media in recent years has fed into some highly undesirable phenomena such as proliferation of offensive language, hate speech, sexist remarks, etc. on the Internet. In light of this, there have been several efforts to automate the detection and moderation of such abusive content. However, deliberate obfuscation of words by users to evade detection poses a serious challenge to the effectiveness of these efforts. The current state of the art approaches to abusive language detection, based on recurrent neural networks, do not explicitly address this problem and resort to a generic OOV (out of vocabulary) embedding for unseen words. However, in using a single embedding for all unseen words we lose the ability to distinguish between obfuscated and non-obfuscated or rare words. In this paper, we address this problem by designing a model that can compose embeddings for…

Author Profiling for Abuse Detection

Pushkar Mishra, Marco Del Tredici, Helen Yannakoudakis +1 more

Artificial Intelligence · UvA-DARE (University of Amsterdam) · · cited 64× · OpenAlex ↗ · more like this ·

The rapid growth of social media in recent years has fed into some highly undesirable phenomena such as proliferation of hateful and offensive language on the Internet. Previous research suggests that such abusive content tends to come from users who share a set of common stereotypes and form communities around them. The current state-of-the-art approaches to abuse detection are oblivious to user and community information and rely entirely on textual (i.e., lexical and semantic) cues. In this paper, we propose a novel approach to this problem that incorporates community-based profiling features of Twitter users. Experimenting with a dataset of 16k tweets, we show that our methods significantly outperform the current state of the art in abuse detection. Further, we conduct a qualitative analysis of model characteristics. We release our code, pre-trained models and all the resources used…

Text-based RL Agents with Commonsense Knowledge: New Challenges, Environments and Baselines

Keerthiram Murugesan, Mattia Atzeni, Pavan Kapanipathi +6 more

cs.AI · · cited 61× · arXiv ↗ · more like this ·

Text-based games have emerged as an important test-bed for Reinforcement Learning (RL) research, requiring RL agents to combine grounded language understanding with sequential decision making. In this paper, we examine the problem of infusing RL agents with commonsense knowledge. Such knowledge would allow agents to efficiently act in the world by pruning out implausible actions, and to perform look-ahead planning to determine how current actions might affect future world states. We design a new text-based gaming environment called TextWorld Commonsense (TWC) for training and evaluating RL agents with a specific kind of commonsense knowledge about objects, their attributes, and affordances. We also introduce several baseline RL agents which track the sequential context and dynamically retrieve the relevant commonsense knowledge from ConceptNet. We show that agents which incorporate…

Joint Associations of Race, Ethnicity, and Socioeconomic Status With Mortality in the Multiethnic Cohort Study

Meera Sangaramoorthy, Salma Shariff‐Marco, Shannon M. Conroy +8 more

Health · JAMA Network Open · · cited 58× · DOI ↗ · more like this ·

Importance: Socioeconomic status may help delineate racial and ethnic inequities in mortality. Objective: To investigate the joint associations of race, ethnicity, and neighborhood and individual socioeconomic status with mortality. Design, Setting, and Participants: This prospective analysis used data from the Multiethnic Cohort Study. A population-based sample of participants recruited from California (mainly Los Angeles County) and Hawaii from 1993 to 1996 was followed up until 2013. African American, European American, Japanese American, Latino American, and Native Hawaiian men and women were included. Participants with baseline residential addresses that could not be geocoded or who were missing information on education or adjustment variables were excluded. Data analyses were conducted from January 2018 to December 2020. Exposures: Neighborhood socioeconomic status (nSES) was…

Explanation and Use of Uncertainty Quantified by Bayesian Neural Network Classifiers for Breast Histopathology Images

Ponkrshnan Thiagarajan, Pushkar Khairnar, Susanta Ghosh

cs.CV · · cited 58× · arXiv ↗ · more like this ·

Despite the promise of Convolutional neural network (CNN) based classification models for histopathological images, it is infeasible to quantify its uncertainties. Moreover, CNNs may suffer from overfitting when the data is biased. We show that Bayesian-CNN can overcome these limitations by regularizing automatically and by quantifying the uncertainty. We have developed a novel technique to utilize the uncertainties provided by the Bayesian-CNN that significantly improves the performance on a large fraction of the test data (about 6% improvement in accuracy on 77% of test data). Further, we provide a novel explanation for the uncertainty by projecting the data into a low dimensional space through a nonlinear dimensionality reduction technique. This dimensionality reduction enables interpretation of the test data through visualization and reveals the structure of the data in a low…

Automatic Cricket Highlight Generation Using Event-Driven and Excitement-Based Features

Pushkar Shukla, Hemant Sadana, Apaar Bansal +4 more

Computer Vision and Pattern Recognition · · cited 57× · DOI ↗ · more like this ·

Producing sports highlights is a labor-intensive work that requires some degree of specialization. We propose a model capable of automatically generating sports highlights with a focus on cricket. Cricket is a sport with a complex set of rules and is played for a longer time than most other sports. In this paper we propose a model that considers both event-based and excitement-based features to recognize and clip important events in a cricket match. Replays, audio intensity, player celebration, and playfield scenarios are examples of cues used to capture such events. To evaluate our framework, we conducted a set of experiments ranging from user studies to a comparison analysis between our highlights and the ones provided by the official broadcasters. The general approval by users and the significant overlap between both kinds of highlights indicate the usefulness of our model in…