Search

Humans as social beings tend to form and identify as groups on ethnic, geographic, religious, ideological, linguistic basis, etc. Once such a group identity is formed, those who identify also demand recognition, sometimes for the purpose of prestige and other times for the sake of political representation. Identity in the modern world, premised on the principles of democracy and freedom, is considered a basic human right. Pakistan is also one such state that recognizes people’s right to association. However, in Pakistan, Gilgit-Baltistan probably is athe only tract of land in comity of nations that, unlike secessionist movements across the globe, has continued to struggle for complete integration with the state of Pakistan. Yet tThe denial of political identity continues even after more than 70 75 years. This study will undertakes a chronological examination of major events and reasons that led to political identity crisis of GB. Federalism as a conflict resolution theory will be, thus, utilized in this study to make a case for G-B’s valid concerns. Henceforth, the political conundrum of Gilgit-Baltistan should reach clarity as further lingering of the conflict will exacerbate into intractable conflictthe overall situation.

Background: Kidney transplantation is the treatment of choice for end stage renal disease (ESRD) patients. It increases life expectancy, improves quality of life, achieves greater physical and occupational rehabilitation compared with maintenance haemodialysis. The rationale of this study is to determine reasons for preference of lifelong haemodialysis over kidney transplantation. This study may give us an insight to how this important and life altering decision is made by the patients. Objective: To determine the reasons for preference of kidney transplantation over lifelong hemodialysis.Material and Methods: A descriptive cross-sectional study on 329 ESRD patients was conducted in Nephrology Department, Khyber Teaching Hospital, Peshawar. During a period of 8 months, from September 2021 till April 2022. Data was collected by non-Probability consecutive sampling technique. Subjects who were willing to participate were included in the study after taking written informed consent.Results: A total of 329 patients were observed with mean age of 51 years with SD ± 10.21. Male patients were 204 (62%), and 125(38%) patients were female. In our study 227(69%) patients agreed for kidney transplant while 102(31%) patients refused kidney transplantation. Main reason for acceptance was freedom from haemodialysis 122(37%), on the other hand, 10(3%) patients refused because of surgical complications of transplant surgery & 33(10%) patients refused because of financial constraints.Conclusion: Our study concludes that 69% of ESRD patients were willing for kidney transplantation, main reasons being freedom from haemodialysis, fluid constraints, diet restrictions.Keywords: Kidney transplant, End stage renal disease (ESRD), Hemodialysis.

Objective: To assess the clinical spectrum and various etiological factors of idiopathic intracranial hypertension in Pakistani patients. Study Design: Cross-sectional study. Place and Duration: Neuro-Ophthalmology Clinic, Armed Forces Institute of Ophthalmology, Rawalpindi Pakistan, from Jan 2017 to Jul 2018. Methodology: A total of 64 individuals with intracranial hypertension (IH) were screened. On the basis of inclusion criteria 32 patients were selected for the study. The demographics, clinical features and etiologies were recorded for each patient. Results: High frequency was noted among females, affecting 26 (84.37%) as compared to males who were 6 (18.75%) in number. Half of the patients had body weight lying in range of overweight, 16 (50%), however, only 5 (15.6%) patients fell into the category of obesity. Among clinical features, headache was the most prominent symptom seen in 17 (53.12%) patients and severe disc swelling seen in 19 (59.3%) was the most common sign. Most prevalent systematic association was anemia, 11 (34.3%), and the most prominent etiology was idiopathic intracranial hypertension, being 26 (81.25%). Conclusion: There was high frequency of intracranial hypertension in the females with obesity, pregnancy and in the age group of 20-30 years.

Objective: To assess the rate of positivity of Toxoplasma gondii, Rubella, Cytomegalovirus, Herpes simplex (TORCH) pathogens in patients suffering from congenital cataract in Pakistan. Study Design: Cross sectional comparative study. Place and Duration of Study: Armed Forces Institute of Ophthalmology, Rawalpindi, from Jun 2017 to Mar 2018. Methodology: Cataract group included sixty children already diagnosed with congenital cataract and control group included 60 children not suffering from congenital cataract. Serum samples of all children were collected to test against IgG and IgM antibodies of Toxoplasma gondii, Rubella, Cytomegalo, Herpes simplex pathogens using enzyme linked immune essay (ELISA) technique. Data were analyzed observe in difference between unilateral and bilateral cataract and between cataract and control group. Results: Out of 60 children in cataract group, 29 were tested positive for Toxoplasma gondii, Rubella, Cytomegalo (CMV), Herpes simplex infections. CMV IgG antibodies were identified in 21 (35%) children followed by Rubella IgG 13 (21.7%), Toxoplasmosis IgG 8 (13.3%), CMV IgM 7 (11.7%), HSV I IgG 5 (8.3%), HSV II IgG 2 (3.3%) and HSV I IgM 1 (1.7%) antibodies. None of the children was positive for Toxoplasmosis, Rubella and HSV II IgM antibodies. There was no statistical difference in rate of positivity of Toxoplasma gondii, Rubella, Cytomegalovirus, Herpes simplex pathogens between unilateral and bilateral cataract and between cataract and control group. Conclusion: Toxoplasma gondii, Rubella, Cytomegalovirus, Herpes simplex pathogens are an important cause of congenital cataract with Cytomegalovirus and Rubella virus being most common respectively. In order to prevent paediatric complications, we need........

Objective: To compare the corneal epithelium healing time and post-operative pain after pterygium surgery with and without he use of autologous serum drops. Study Design: Comparative cross-sectional study. Place and Duration of Study: Eye Department, Combined Military Hospital Quetta, from Jan to Aug 2019. Methodology: Twenty patients were included in this study and divided into two groups. All patients underwent pterygium excision with conjunctival autograft. Fourteen of them were instilled with autologous serum drops while 14 were given lubricants postoperatively until corneal epithelium healed completely. Corneal epithelium healing time, visual analog scale for postoperative pain assessment, conjunctival inflammation and recurrences were evaluated. Patients were followed up for 6 months. Results: For corneal epithelial healing, 1 (7.14%) patient showed complete healing on day 2 while 7 (50%) showed healing on day 3 and 6 (42.86%) on day 4. While in non-autologous serum drops group 7 (50%) showed complete healing on day 4 while 4 (28.57%) showed healing on day 5 and 3 (21.43%) on day 6. The mean corneal healing time of autologous serum drops group was 3.36 ± 0.63 days and that of non-autologous serum drops group was 4.71 ± 0.83 days. Patients of autologous serum drops group also experienced significantly less pain as compared to the non-autologous serum drops group. While the recurrence rate and conjunctival inflammation in the two groups was found to be of equal degree. Conclusion: Patients who used autologous serum eye drops post operatively experienced less pain and their corneal epithelium healed earlier than the.......

After the menopause the female is unable to imitate further. The most important factor in menopause is the aging of female. Females usually can reproduce up to 50 years. Many indications appeared in the average age of the female are related to menopausal evolution. It may cause the more severe symptoms. It causes disorder of aging by creating bone weakness and heart issues. The destruction of follicles and presence of different lighter and severe organic and psychosomatic shriek are under investigation. Reduction of performance scheme based on significant, consistent and intention conditions are the biggest hurdle in the development of investigation on reproductive aging and malfunction of ovaries. Naming system used recently is defined and its boundaries are also conferred. The female’s reproductive system cannot be Described by modern vocabulary. Because there is a reduction of compassion and specificity in the modern vocabulary. For the recognition of suitable differentiation throughout the life time of reproduction, various staging standards are estimated. More investigations will be helpful in more understanding of menopausal evolution. And these are also essential to determine the strength, realism and suitability of planned performance standards.The permanent malfunction of ovaries is called menopause. Place and duration: In the Obstetrics and Gynecology department of Mayo Hospital Lahore for one-year duration from December 2017 to November 2018. Keywords: Menopause; Menopausal transition; per menopause; Postmenopausal

Objective: To study the spectrum of optic neuropathies in patient’s presenting at AFIO Rawalpindi. Study Design: Descriptive study/retrospective study. Place and Duration of Study: Armed Forces Institute of Ophthalmology, Rawalpindi, from Jan 2017 to Dec 2017. Methodology: Retrospective analysis of hospital record of patients coming to neuro-ophthalmic clinic of ArmedForces Institute of Ophthalmology (AFIO), Rawalpindi, from Jan to Dec 2017 was done. Only patients with clearfinal diagnosis of neurophthalmic disease were included. Those with vague diagnosis were excluded. Data wasentered in SPSS version 22 and grouped into variables before analysis. Chi-square test was used for statisticalanalysis. The p

The main purpose of this work is to solve the problem of electromagnetic scattering from a chiral cylinder of arbitrary cross-section above an infinite perfect electric conducting (PEC) plane. Using image theory, this problem is reduced to two chiral cylinders in free-space. Surface equivalence principle is used to obtain three different equivalent problems for this two-cylinder problem. Then, the method of moments is used to solve these equivalent problems numerically. It is known that the image of a chiral body through a ground plane is another chiral body with the same permittivity and permeability but opposite chirality. Using this property, the two-body problem in the moment method may be reduced to a one-body problem with a complicated moment matrix. Computed numerical results include scattered fields and equivalent surface currents on the cylinder. Turkish Academy of Sciences

Objective: To compare the difference in pain perceived by patients undergoing intra-oral local anesthesia withdifferent gauge needles. Study Design: Qausi experimental study. Place and Duration of Study: Oral and Maxillofacial surgery department of Institute of Dentistry, CombinedMilitary Hospital, Lahore Medical College, Lahore, from July 2019 to August 2019. Methodology: One hundred patients were selected by consecutive sampling from the oral surgery OPD clinic.They were divided into two groups A and B randomly. Twenty three gauge needle on a 3cc disposable syringewas used for inferior alveolar nerve block and buccal nerve block for group A and 27 gauge needle on a metaldental syringe was used for the same in group B. Patients gave a verbal pain score, from 0-10 for each injection. Results: One Hundred participants were included in study, 37 (37%) males and 63 (63%) females. Mean painscore for group A for the inferior alveolar nerve block was 4.50 ± 2.1 and group B was 3.86 ± 2. The mean painscore for the buccal nerve block in group A was 4.02 ± 1.7, while that of group B was 3.94 ± 1.8. There was nosignificant difference (p=0.167 & 0.855) in pain perceived by patients undergoing intra oral local anesthesia using needles of different gauges. Conclusion: There is no difference in pain perceived by patients undergoing intra oral local anesthesia usingneedles of different gauges.

Background: While it has been shown that aerobic exercise interventions are well tolerated in participants with the Huntington disease (HD) gene mutation, no study to date has tested whether an aerobic exercise intervention benefits brain structure and function in pre-manifest HD. Objective: In this study we utilized magnetic resonance (MR) imaging techniques to assess the efficacy of moderate-to-vigorous exercise treatment relative to active stretching and toning control. Methods: Forty pre-manifest participants with confirmed HD gene expansion were recruited into a two-arm intervention study that included a moderate-to-vigorous intensity home-based walking exercise intervention (N=34) and an active stretching and toning control intervention (N=6). Participants were assessed at baseline and after 26 weeks in one of the two study arms. Results: 25 of the 34 (74%) participants assigned to the moderate-to-vigorous intensity group completed the intervention while 4 of the 6 (67%) participants in the stretching and toning intervention completed the study. The primary analyses compared the two arms of the study and found no statistical differences between the groups. Both groups were found to have improved their cardiorespiratory fitness as assessed by maximal oxygen uptake (VO2max). A secondary analysis combined the two arms of the study and there was a significant relationship (p

A dual-band wearable antenna is designed on a textile material. The design operates at ISM bands available for Wireless Body Area Network and Wireless Local Area Network with an input match better than -15 dB. The antenna is designed by using Computational Electromagnetic Software (CEMS) based on finite-difference time-domain (FDTD) method. A three-layer phantom model including skin, fat and muscle has been considered to compute the specific absorption rate (SAR). The maximum value of SAR averaged over 1g and 10g of tissue is less than 1.6 W/Kg and 2 W/Kg, respectively, when the maximum incident power of the antenna is 63 mW. These values are incompliance with the international electromagnetic safety standards.

Transfer learning from pre-trained neural language models towards downstream tasks has been a predominant theme in NLP recently. Several researchers have shown that deep NLP models learn non-trivial amount of linguistic knowledge, captured at different layers of the model. We investigate how fine-tuning towards downstream NLP tasks impacts the learned linguistic knowledge. We carry out a study across popular pre-trained models BERT, RoBERTa and XLNet using layer and neuron-level diagnostic classifiers. We found that for some GLUE tasks, the network relies on the core linguistic information and preserve it deeper in the network, while for others it forgets. Linguistic information is distributed in the pre-trained language models but becomes localized to the lower layers post fine-tuning, reserving higher layers for the task specific knowledge. The pattern varies across architectures, with BERT retaining linguistic information relatively deeper in the network compared to RoBERTa and XLNet, where it is predominantly delegated to the lower layers., Comment: Findings of the ACL 2021

This paper is a write-up for the tutorial on "Fine-grained Interpretation and Causation Analysis in Deep NLP Models" that we are presenting at NAACL 2021. We present and discuss the research work on interpreting fine-grained components of a model from two perspectives, i) fine-grained interpretation, ii) causation analysis. The former introduces methods to analyze individual neurons and a group of neurons with respect to a language property or a task. The latter studies the role of neurons and input features in explaining decisions made by the model. We also discuss application of neuron analysis such as network manipulation and domain adaptation. Moreover, we present two toolkits namely NeuroX and Captum, that support functionalities discussed in this tutorial., Comment: Accepted at NAACL Tutorial

While a lot of analysis has been carried to demonstrate linguistic knowledge captured by the representations learned within deep NLP models, very little attention has been paid towards individual neurons.We carry outa neuron-level analysis using core linguistic tasks of predicting morphology, syntax and semantics, on pre-trained language models, with questions like: i) do individual neurons in pre-trained models capture linguistic information? ii) which parts of the network learn more about certain linguistic phenomena? iii) how distributed or focused is the information? and iv) how do various architectures differ in learning these properties? We found small subsets of neurons to predict linguistic tasks, with lower level tasks (such as morphology) localized in fewer neurons, compared to higher level task of predicting syntax. Our study also reveals interesting cross architectural comparisons. For example, we found neurons in XLNet to be more localized and disjoint when predicting properties compared to BERT and others, where they are more distributed and coupled., Accepted in EMNLP 2020

With the outbreak of the COVID-19 pandemic, people turned to social media to read and to share timely information including statistics, warnings, advice, and inspirational stories. Unfortunately, alongside all this useful information, there was also a new blending of medical and political misinformation and disinformation, which gave rise to the first global infodemic. While fighting this infodemic is typically thought of in terms of factuality, the problem is much broader as malicious content includes not only fake news, rumors, and conspiracy theories, but also promotion of fake cures, panic, racism, xenophobia, and mistrust in the authorities, among others. This is a complex problem that needs a holistic approach combining the perspectives of journalists, fact-checkers, policymakers, government entities, social media platforms, and society as a whole. Taking them into account we define an annotation schema and detailed annotation instructions, which reflect these perspectives. We performed initial annotations using this schema, and our initial experiments demonstrated sizable improvements over the baselines. Now, we issue a call to arms to the research community and beyond to join the fight by supporting our crowdsourcing annotation efforts., COVID-19, Infodemic, Disinformation, Misinformation, Fake News, Call to Arms, Crowdsourcing Annotations

Mutual coupling compensation in uniform linear and circular receiving antenna arrays of thin wire dipoles is presented. It was observed that the mutual impedance is independent of the incident angle and depends solely on the geometry of the array. By using only one measurement, decoupling matrix is computed and direction of arrival is estimated.

Background and Aims The patients of end stage renal disease suffer a lot of psychological trauma as well as feel a lot of social burden. They mostly show psychological distress in different forms, be it impaired quality of life, depression or anxiety. The aim of our study was to find out the effect of supportive-expressive group therapy in end stage kidney disease patients in the outcome of better health related quality of life and survival. Method 181 patients were recruited in the study after consent between June 2012-June 2017. They were then assigned randomly to Group A supportive-expressive group therapy or to a Group B control group (no intervention). It was a double blinded randomized control trial. 115 patients were assigned to Group A and 66 were assigned to group B. The primary outcome was the effect on health related quality of life. Secondly, if it had any impact on survival. Significant predictors of survival were not analysed. Analysis was done by intention to treat. Results The patients of group A showed decrease in physical pain(p=0.002), mood elevation (p=0.04), compliance with treatment (p=0.001), reduced hopelessness (p=0.004), newly diagnosed depressive disorders (p=0.01) and better social life (p=0.003). Group A had 97% survival rate as compared to group B with 77%. Conclusion Supportive-expressive group therapy to patients with end stage kidney disease improves health related quality of life. It also has an effect on survival maybe because of compliance with treatment. We would like to further explore the role of Supportive-expressive group therapy in survival. Support groups should be provided to patients in developing countries like the setup of ours because it has a great impact on the betterment of the patient.

Roman Urdu is an informal form of the Urdu language written in Roman script, which is widely used in South Asia for online textual content. It lacks standard spelling and hence poses several normalization challenges during automatic language processing. In this article, we present a feature-based clustering framework for the lexical normalization of Roman Urdu corpora, which includes a phonetic algorithm UrduPhone, a string matching component, a feature-based similarity function, and a clustering algorithm Lex-Var. UrduPhone encodes Roman Urdu strings to their pronunciation-based representations. The string matching component handles character-level variations that occur when writing Urdu using Roman script. The similarity function incorporates various phonetic-based, string-based, and contextual features of words. The Lex-Var algorithm is a variant of the k-medoids clustering algorithm that groups lexical variations of words. It contains a similarity threshold to balance the number of clusters and their maximum similarity. The framework allows feature learning and optimization in addition to the use of predefined features and weights. We evaluate our framework extensively on four real-world datasets and show an F-measure gain of up to 15% from baseline methods. We also demonstrate the superiority of UrduPhone and Lex-Var in comparison to respective alternate algorithms in our clustering framework for the lexical normalization of Roman Urdu.

Pre-trained Transformer-based models have achieved state-of-the-art performance for various Natural Language Processing (NLP) tasks. However, these models often have billions of parameters, and, thus, are too resource-hungry and computation-intensive to suit low-capability devices or applications with strict latency requirements. One potential remedy for this is model compression, which has attracted a lot of research attention. Here, we summarize the research in compressing Transformers, focusing on the especially popular BERT model. In particular, we survey the state of the art in compression for BERT, we clarify the current best practices for compressing large-scale Transformer models, and we provide insights into the workings of various methods. Our categorization and analysis also shed light on promising future research directions for achieving lightweight, accurate, and generic NLP models., To appear in TACL 2021. The arXiv version is a pre-MIT Press publication version

Transformer-based deep NLP models are trained using hundreds of millions of parameters, limiting their applicability in computationally constrained environments. In this paper, we study the cause of these limitations by defining a notion of Redundancy, which we categorize into two classes: General Redundancy and Task-specific Redundancy. We dissect two popular pretrained models, BERT and XLNet, studying how much redundancy they exhibit at a representation-level and at a more fine-grained neuron-level. Our analysis reveals interesting insights, such as: i) 85% of the neurons across the network are redundant and ii) at least 92% of them can be removed when optimizing towards a downstream task. Based on our analysis, we present an efficient feature-based transfer learning procedure, which maintains 97% performance while using at-most 10% of the original neurons., 19 Pages, 14 figures, EMNLP 2020

The widespread usage of Twitter during emergencies has provided a new opportunity and timely resource to crisis responders for various disaster management tasks. Geolocation information of pertinent tweets is crucial for gaining situational awareness and delivering aid. However, the majority of tweets do not come with geoinformation. In this work, we focus on the task of location mention recognition from crisis-related tweets. Specifically, we investigate the influence of different types of labeled training data on the performance of a BERT-based classification model. We explore several training settings such as combing in- and out-domain data from news articles and general-purpose and crisis-related tweets. Furthermore, we investigate the effect of geospatial proximity while training on near or far-away events from the target event. Using five different datasets, our extensive experiments provide answers to several critical research questions that are useful for the research community to foster research in this important direction. For example, results show that, for training a location mention recognition model, Twitter-based data is preferred over general-purpose data; and crisis-related data is preferred over general-purpose Twitter data. Furthermore, training on data from geographically-nearby disaster events to the target event boosts the performance compared to training on distant events.

Transformer-based NLP models are trained using hundreds of millions or even billions of parameters, limiting their applicability in computationally constrained environments. While the number of parameters generally correlates with performance, it is not clear whether the entire network is required for a downstream task. Motivated by the recent work on pruning and distilling pre-trained models, we explore strategies to drop layers in pre-trained models, and observe the effect of pruning on downstream GLUE tasks. We were able to prune BERT, RoBERTa and XLNet models up to 40%, while maintaining up to 98% of their original performance. Additionally we show that our pruned models are on par with those built using knowledge distillation, both in terms of size and performance. Our experiments yield interesting observations such as, (i) the lower layers are most critical to maintain downstream task performance, (ii) some tasks such as paraphrase detection and sentence similarity are more robust to the dropping of layers, and (iii) models trained using a different objective function exhibit different learning patterns and w.r.t the layer dropping.

Mutual coupling is compensated in a four element uniform linear receiving array using software defined radios. Direction of arrival (DoA) is estimated in real-time for the array with spacing d=lambda/4. The decoupling matrix was measured using a VNA for only one incident angle. After compensation the error in DoA estimation was reduced to 5%. Comparing the DoA results with d=lambda/2 spaced Uniform Linear Array (ULA), 1.2% error was observed. Although, the experiment was performed indoors with a low SNR, the results show a substantial improvement in the estimated DoA after compensation. Türkiye Bilimsel ve Teknolojik Araştırma Kurumu (TUBITAK)

Two sets of novel extensions of NNJM model are proposed. The NDAM models that regularizes the loss function with respect to in-domain model, give an improvement of up to +0.4 BLEU points. The NFM models that fuse in- and out-domain NNJM models give an improvement of up to +0.9 BLEU points. The NFM models also beat state-of-the-art phrase-table adaptation methods. The gains obtained from NNJM and phrase-table adaptation were found to be additive. We explore neural joint models for the task of domain adaptation in machine translation in two ways: (i)we apply state-of-the-art domain adaptation techniques, such as mixture modelling and data selection using the recently proposed Neural Network Joint Model (NNJM) (Devlin etal., 2014); (ii)we propose two novel approaches to perform adaptation through instance weighting and weight readjustment in the NNJM framework. In our first approach, we propose a pair of models called Neural Domain Adaptation Models (NDAM) that minimizes the cross entropy by regularizing the loss function with respect to in-domain (and optionally to out-domain) model. In the second approach, we present a set of Neural Fusion Models (NFM) that combines the in- and the out-domain models by readjusting their parameters based on the in-domain data.We evaluated our models on the standard task of translating English-to-German and Arabic-to-English TED talks. The NDAM models achieved better perplexities and modest BLEU improvements compared to the baseline NNJM, trained either on in-domain or on a concatenation of in- and out-domain data. On the other hand, the NFM models obtained significant improvements of up to +0.9 and +0.7 BLEU points, respectively. We also demonstrate improvements over existing adaptation methods such as instance weighting, phrasetable fill-up, linear and log-linear interpolations.

The role of social media, in particular microblogging platforms such as Twitter, as a conduit for actionable and tactical information during disasters is increasingly acknowledged. However, time-critical analysis of big crisis data on social media streams brings challenges to machine learning techniques, especially the ones that use supervised learning. The scarcity of labeled data, particularly in the early hours of a crisis, delays the learning process. Existing classification methods require a significant amount of labeled data specific to a particular event for training plus a lot of feature engineering to achieve best results. In this work, we introduce neural network based classification methods for identifying useful tweets during a crisis situation. At the onset of a disaster when no labeled data is available, our proposed method makes the best use of the out-of-event data and achieves good results.

We share the findings of the first shared task on improving robustness of Machine Translation (MT). The task provides a testbed representing challenges facing MT models deployed in the real world, and facilitates new approaches to improve models; robustness to noisy input and domain mismatch. We focus on two language pairs (English-French and English-Japanese), and the submitted systems are evaluated on a blind test set consisting of noisy comments on Reddit and professionally sourced translations. As a new task, we received 23 submissions by 11 participating teams from universities, companies, national labs, etc. All submitted systems achieved large improvements over baselines, with the best improvement having +22.33 BLEU. We evaluated submissions by both human judgment and automatic evaluation (BLEU), which shows high correlations (Pearson's r = 0.94 and 0.95). Furthermore, we conducted a qualitative analysis of the submitted systems using compare-mt, which revealed their salient differences in handling challenges in this task. Such analysis provides additional insights when there is occasional disagreement between human judgment and BLEU, e.g. systems better at producing colloquial expressions received higher score from human judgment.

We address the problem of simultaneous translation by modifying the Neural MT decoder to operate with dynamically built encoder and attention. We propose a tunable agent which decides the best segmentation strategy for a user-defined BLEU loss and Average Proportion (AP) constraint. Our agent outperforms previously proposed Wait-if-diff and Wait-if-worse agents (Cho and Esipova, 2016) on BLEU with a lower latency. Secondly we proposed data-driven changes to Neural MT training to better match the incremental decoding framework., NAACL 2018

A 1:16 Wilkinson power divider (WPD) working from 1.3 GHz to 3 GHz is presented in this paper. Equal power division, phase and amplitude consistency is achieved among the output ports. The proposed design has a wideband with a 10 dB fractional bandwidth of 76%. The isolation loss within the bandwidth is better than 15 dB and the insertion loss is less than 2 dB. The size of the WPD has been minimized to 16×10,7 cm 2, which was the main objective of this work. The compact size of the proposed design makes it a suitable candidate for feeding phased antenna arrays for wideband operations. IEEE IEEE Karachi Section Pakistan Science Foundation Institution of Engineers, Pakistan (IEP) Hamdard Foundation

Despite the remarkable evolution of deep neural networks in natural language processing (NLP), their interpretability remains a challenge. Previous work largely focused on what these models learn at the representation level. We break this analysis down further and study individual dimensions (neurons) in the vector representation learned by end-to-end neural models in NLP tasks. We propose two methods: Linguistic Correlation Analysis, based on a supervised method to extract the most relevant neurons with respect to an extrinsic task, and Cross-model Correlation Analysis, an unsupervised method to extract salient neurons w.r.t. the model itself. We evaluate the effectiveness of our techniques by ablating the identified neurons and reevaluating the network's performance for two tasks: neural machine translation (NMT) and neural language modeling (NLM). We further present a comprehensive analysis of neurons with the aim to address the following questions: i) how localized or distributed are different linguistic properties in the models? ii) are certain neurons exclusive to some properties and not others? iii) is the information more or less distributed in NMT vs. NLM? and iv) how important are the neurons identified through the linguistic correlation method to the overall task? Our code is publicly available as part of the NeuroX toolkit (Dalvi et al. 2019)., Comment: AAA 2019, pages 10, AAAI Conference on Artificial Intelligence (AAAI 2019)

We present a toolkit to facilitate the interpretation and understanding of neural network models. The toolkit provides several methods to identify salient neurons with respect to the model itself or an external task. A user can visualize selected neurons, ablate them to measure their effect on the model accuracy, and manipulate them to control the behavior of the model at the test time. Such an analysis has a potential to serve as a springboard in various research directions, such as understanding the model, better architectural choices, model distillation and controlling data biases., Comment: AAAI Conference on Artificial Intelligence (AAAI 2019), Demonstration track, pages 2