Search

We experimentally study III-V/Ge heterostructure and demonstrate InGaAs hetero-contacts to n-Ge with a wide range of In % and achieve low contact resistivity ($\rho_C$) of $5\times10^{-8} \Omega\cdot cm^2$ for Ge doping of $3 \times 10^{19} cm^{-3}$. This results from re-directing the charge neutrality level (CNL) near the conduction band and benefiting from low effective mass for high electron transmission. For the first time, we observe that the heterointerface presents no temperature dependence despite the two different conduction minimum valley locations of III-V ($\Gamma$-valley) and Ge (L-valley), which potentially stems from elastic trap-assisted tunneling through defect states at the interface generated by dislocations. The hetero-interface plays a dominant role in the overall $\rho_C$ below $\approx 1 \times 10^{-7} \Omega \cdot cm^2$, which can be further improved with large active dopant concentration in Ge by co-doping., Comment: 2 pages, 16 figures; Accepted in SSDM 2017

The Internet is responsible for accelerating growth in several fields such as digital media, healthcare, the military. Furthermore, the Internet was founded on the principle of allowing clients to communicating with servers. However, serverless computing is one such field that tries to break free from this paradigm. Event-driven compute services allow users to build more agile applications using capacity provisioning and a pay-for-value billing model. This paper provides a formal account of the research contributions in the field of Serverless computing., Comment: 8 pages, 1 figure

In order to understand the resourcefulness of a natural quantum system in quantum communication tasks, we study the dense coding capacity (DCC) and teleportation fidelity (TF) of Haar uniformly generated random multipartite states of various ranks. We prove that when a rank-2 two-qubit state, a Werner state, and a pure state possess the same amount of entanglement, the DCC of a rank-2 state belongs to the envelope made by pure and Werner states. In a similar way, we obtain an upper bound via the generalized Greenberger-Horne-Zeilinger state for rank-2 three-qubit states when the dense coding with two senders and a single receiver is performed and entanglement is measured in the senders:receiver bipartition. The normalized frequency distribution of DCC for randomly generated two-, three- and four-qubit density matrices with global as well as local decodings at the receiver's end are reported. The estimation of mean DCC for two-qubit states is found to be in good agreement with the numerical simulations. Universally, we observe that the performance of random states for dense coding as well as teleportation decreases with the increase of the rank of states which we have shown to be surmounted by the local pre-processing operations performed on the shared states before starting the protocols, irrespective of the rank of the states. The local pre-processing employed here is based on positive operator valued measurements along with classical communication and we show that unlike dense coding with two-qubit random states, senders' operations are always helpful to probabilistically enhance the capabilities of implementing dense coding as well as teleportation., Comment: V1: 15 pages, 9 figures; V2: Close to published version, updated figures and tables

We show that a perfectly genuine tripartite EPR steerable assemblage can be distilled from partially genuine tripartite EPR steerable assemblages. In particular, we consider two types of hybrid scenarios: one-sided device-independent (1SDI) scenario (where one observer is untrusted, and other two observers are trusted) and two-sided device-independent (2SDI) scenario (where two observers are untrusted, and one observer is trusted). In both the scenarios, we show distillation of perfectly genuine steerable assemblage of three-qubit Greenberger-Horne-Zeilinger (GHZ) states or three-qubit W states from many copies of initial partially genuine steerable assemblages of the corresponding states. In each of these cases, we demonstrate that at least one copy of a perfectly genuine steerable assemblage can be distilled with certainty from infinitely many copies of initial assemblages. In case of practical scenarios employing finite copies, we show that the efficiency of our distillation protocols reaches near perfect levels using only a few number of initial assemblages., Comment: Close to the published version

Combining symbolic human knowledge with neural networks provides a rule-based ante-hoc explanation of the output. In this paper, we propose feature extracting functions for integrating human knowledge abstracted as logic rules into the predictive behavior of a neural network. These functions are embodied as programming functions, which represent the applicable domain knowledge as a set of logical instructions and provide a modified distribution of independent features on input data. Unlike other existing neural logic approaches, the programmatic nature of these functions implies that they do not require any kind of special mathematical encoding, which makes our method very general and flexible in nature. We illustrate the performance of our approach for sentiment classification and compare our results to those obtained using two baselines., Comment: to be published in S+SSPR 2020 proceedings

This paper describes our efforts in tackling Task 5 of SemEval-2020. The task involved detecting a class of textual expressions known as counterfactuals and separating them into their constituent elements. Counterfactual statements describe events that have not or could not have occurred and the possible implications of such events. While counterfactual reasoning is natural for humans, understanding these expressions is difficult for artificial agents due to a variety of linguistic subtleties. Our final submitted approaches were an ensemble of various fine-tuned transformer-based and CNN-based models for the first subtask and a transformer model with dependency tree information for the second subtask. We ranked 4-th and 9-th in the overall leaderboard. We also explored various other approaches that involved the use of classical methods, other neural architectures and the incorporation of different linguistic features., Comment: 10 pages, 1 figure, 4 tables. For associated code, see https://github.com/gargrohin/Counterfactuals-NLP. Accepted at Proceedings of 14th International Workshop on Semantic Evaluation (SemEval-2020)

We investigate the possibility of multiple use of a single copy of three-qubit states for genuine tripartite Einstein-Podolsky-Rosen (EPR) steering. A pure three-qubit state of either the Greenberger-Horne-Zeilinger (GHZ)-type or W-type is shared between two fixed observers in two wings and a sequence of multiple observers in the third wing who perform unsharp or non-projective measurements. The choice of measurement settings for each of the multiple observers in the third wing is independent and uncorrelated with the measurement settings and outcomes of the previous observers. We investigate all possible types of (2->1) and (1->2) genuine tripartite steering in the above set-up. For each case, we obtain an upper limit on the number of observers on the third wing who can demonstrate genuine EPR steering through violation of a tripartite steering inequality. We show that the GHZ state allows for a higher number of observers compared to that for W state. Additionally, (1->2) steering is possible for a larger range of the sharpness parameter compared to that for the (2->1) steering cases., Comment: Close to the published version

We explore the impact of non-Markovian channels on the quantum correlations (QCs) of Haar uniformly generated random two-qubit input states with different ranks -- either one of the qubits (single-sided) or both the qubits independently (double-sided) are passed through noisy channels. Under dephasing and depolarizing channels with varying non-Markovian strength, entanglement and quantum discord of the output states collapse and revive with the increase of noise. We find that in case of the depolarizing double-sided channel, both the QCs of random states show a higher number of revivals on average than that of the single-sided ones with a fixed non-Markovianity strength, irrespective of the rank of the states -- we call such a counter-intuitive event as a constructive feedback of non-Markovianity. Consequently, the average noise at which QCs of random states show first revival decreases with the increase of the strength of non-Markovian noise, thereby indicating the role of non-Markovian channels on the regenerations of QCs even in presence of a high amount of noise. However, we observe that non-Markovianity does not play any role to increase the robustness in random quantum states which can be measured by the mean value of critical noise at which quantum correlations first collapse. Moreover, we observe that the tendency of a state to show regeneration increases with the increase of average QCs of the random input states along with non-Markovianity., Comment: V3: Close to published version

We consider the task of certification of genuine entanglement of tripartite states. For this purpose, we first present an "all-versus-nothing" proof of genuine tripartite Einstein-Podolsky-Rosen (EPR) steering by demonstrating the non-existence of a hybrid local hidden state (LHS) model in the tripartite network as a motivation to our main result. A full logical contradiction of the predictions of the hybrid LHS model with quantum mechanical outcome statistics for any three-qubit generalized Greenberger-Horne-Zeilinger (GGHZ) states and pure W-class states is shown. Using logical contradiction, we can distinguish between the GGHZ and W-class state in a two-sided device-independent (2SDI) steering scenario. We next formulate a 2SDI steering inequality which is a generalization of the fine-grained steering inequality (FGI) derived in \cite{PKM14} for the tripartite scenario. We show that the maximum quantum violation of this tripartite FGI can be used to certify genuine entanglement of three-qubit pure states., Comment: Close to the published version

Detecting whether a news article is fake or genuine is a crucial task in today's digital world where it's easy to create and spread a misleading news article. This is especially true of news stories shared on social media since they don't undergo any stringent journalistic checking associated with main stream media. Given the inherent human tendency to share information with their social connections at a mouse-click, fake news articles masquerading as real ones, tend to spread widely and virally. The presence of echo chambers (people sharing same beliefs) in social networks, only adds to this problem of wide-spread existence of fake news on social media. In this paper, we tackle the problem of fake news detection from social media by exploiting the very presence of echo chambers that exist within the social network of users to obtain an efficient and informative latent representation of the news article. By modeling the echo-chambers as closely-connected communities within the social network, we represent a news article as a 3-mode tensor of the structure - and propose a tensor factorization based method to encode the news article in a latent embedding space preserving the community structure. We also propose an extension of the above method, which jointly models the community and content information of the news article through a coupled matrix-tensor factorization framework. We empirically demonstrate the efficacy of our method for the task of Fake News Detection over two real-world datasets. Further, we validate the generalization of the resulting embeddings over two other auxiliary tasks, namely: \textbf{1)} News Cohort Analysis and \textbf{2)} Collaborative News Recommendation. Our proposed method outperforms appropriate baselines for both the tasks, establishing its generalization., Comment: Presented at ASONAM'18

A tripartite quantum network is said to be bilocal if two independent sources produce a pair of bipartite entangled states. Quantum non-bilocal correlation emerges when the central party which possesses two particles from two different sources performs Bell-state measurement on them and nonlocality is generated between the other two uncorrelated systems in this entanglement-swapping protocol. The interaction of such systems with the environment reduces quantum non-bilocal correlations. Here we show that the diminishing effect modelled by the amplitude damping channel can be slowed down by employing the technique of weak measurements and reversals. It is demonstrated that for a large range of parameters the quantum non-bilocal correlations are preserved against decoherence by taking into account the average success rate of the post-selection governing weak measurements., Comment: 12 pages, 16 figures

Adverse drug reactions (ADRs) are one of the leading causes of mortality in health care. Current ADR surveillance systems are often associated with a substantial time lag before such events are officially published. On the other hand, online social media such as Twitter contain information about ADR events in real-time, much before any official reporting. Current state-of-the-art in ADR mention extraction uses Recurrent Neural Networks (RNN), which typically need large labeled corpora. Towards this end, we propose a multi-task learning based method which can utilize a similar auxiliary task (adverse drug event detection) to enhance the performance of the main task, i.e., ADR extraction. Furthermore, in the absence of auxiliary task dataset, we propose a novel joint multi-task learning method to automatically generate weak supervision dataset for the auxiliary task when a large pool of unlabeled tweets is available. Experiments with 0.48M tweets show that the proposed approach outperforms the state-of-the-art methods for the ADR mention extraction task by 7.2% in terms of F1 score., Comment: Accepted at ECIR18 as full paper (12 pages)

Adverse drug reactions (ADRs) are one of the leading causes of mortality in health care. Current ADR surveillance systems are often associated with a substantial time lag before such events are officially published. On the other hand, online social media such as Twitter contain information about ADR events in real-time, much before any official reporting. Current state-of-the-art methods in ADR mention extraction use Recurrent Neural Networks (RNN), which typically need large labeled corpora. Towards this end, we propose a semi-supervised method based on co-training which can exploit a large pool of unlabeled tweets to augment the limited supervised training data, and as a result enhance the performance. Experiments with 0.1M tweets show that the proposed approach outperforms the state-of-the-art methods for the ADR mention extraction task by 5% in terms of F1 score., Comment: Accepted at ECIR18 as short paper (6 pages)

Social media is an useful platform to share health-related information due to its vast reach. This makes it a good candidate for public-health monitoring tasks, specifically for pharmacovigilance. We study the problem of extraction of Adverse-Drug-Reaction (ADR) mentions from social media, particularly from twitter. Medical information extraction from social media is challenging, mainly due to short and highly information nature of text, as compared to more technical and formal medical reports. Current methods in ADR mention extraction relies on supervised learning methods, which suffers from labeled data scarcity problem. The State-of-the-art method uses deep neural networks, specifically a class of Recurrent Neural Network (RNN) which are Long-Short-Term-Memory networks (LSTMs) \cite{hochreiter1997long}. Deep neural networks, due to their large number of free parameters relies heavily on large annotated corpora for learning the end task. But in real-world, it is hard to get large labeled data, mainly due to heavy cost associated with manual annotation. Towards this end, we propose a novel semi-supervised learning based RNN model, which can leverage unlabeled data also present in abundance on social media. Through experiments we demonstrate the effectiveness of our method, achieving state-of-the-art performance in ADR mention extraction., Comment: Accepted at DTMBIO workshop, CIKM 2017. To appear in BMC Bioinformatics. Pls cite that version

The integration of efficient, miniaturized group IV lasers into CMOS architecture holds the key to the realization of fully functional photonic-integrated circuits. Despite several years of progress, however, all group IV lasers reported to date exhibit impractically high thresholds owing to their unfavorable bandstructures. Highly strained germanium with its fundamentally altered bandstructure has emerged as a potential low-threshold gain medium, but there has yet to be any successful demonstration of lasing from this seemingly promising material system. Here, we demonstrate a low-threshold, compact group IV laser that employs germanium nanowire under a 1.6% uniaxial tensile strain as the gain medium. The amplified material gain in strained germanium can sufficiently surmount optical losses at 83 K, thus allowing the first observation of multimode lasing with an optical pumping threshold density of ~3.0 kW cm^-^2. Our demonstration opens up a new horizon of group IV lasers for photonic-integrated circuits., Comment: 31 pages, 9 figures

Inferring trust relations between social media users is critical for a number of applications wherein users seek credible information. The fact that available trust relations are scarce and skewed makes trust prediction a challenging task. To the best of our knowledge, this is the first work on exploring representation learning for trust prediction. We propose an approach that uses only a small amount of binary user-user trust relations to simultaneously learn user embeddings and a model to predict trust between user pairs. We empirically demonstrate that for trust prediction, our approach outperforms classifier-based approaches which use state-of-the-art representation learning methods like DeepWalk and LINE as features. We also conduct experiments which use embeddings pre-trained with DeepWalk and LINE each as an input to our model, resulting in further performance improvement. Experiments with a dataset of $\sim$356K user pairs show that the proposed method can obtain an high F-score of 92.65%., Comment: To appear in the proceedings of ASONAM'17. Please cite that version

Hate speech detection on Twitter is critical for applications like controversial event extraction, building AI chatterbots, content recommendation, and sentiment analysis. We define this task as being able to classify a tweet as racist, sexist or neither. The complexity of the natural language constructs makes this task very challenging. We perform extensive experiments with multiple deep learning architectures to learn semantic word embeddings to handle this complexity. Our experiments on a benchmark dataset of 16K annotated tweets show that such deep learning methods outperform state-of-the-art char/word n-gram methods by ~18 F1 points., Comment: In Proceedings of ACM WWW'17 Companion, Perth, Western Australia, Apr 2017 (WWW'17), 2 pages