arXiv subjects

Divyanshu Gupta

Publications and source records attributed to Divyanshu Gupta.

6 recordsLinked to original sources

arXiv2024

Understanding Student Sentiment on Mental Health Support in Colleges Using Large Language Models

Mental health support in colleges is vital in educating students by offering counseling services and organizing supportive events. However, evaluating its effectiveness faces challenges like data collection difficulties and lack of standardized metrics, limiting research scope. Student feedback is crucial for evaluation but often relies on qualitative analysis without systematic investigation using advanced machine learning methods. This paper uses public Student Voice Survey data to analyze student sentiments on mental health support with large language models (LLMs). We created a sentiment analysis dataset, SMILE-College, with human-machine collaboration. The investigation of both traditional machine learning methods and state-of-the-art LLMs showed the best performance of GPT-3.5 and BERT on this new dataset. The analysis highlights challenges in accurately predicting response sentiments and offers practical insights on how LLMs can enhance mental health-related research and improve college mental health services. This data-driven approach will facilitate efficient and informed mental health support evaluation, management, and decision-making.

cs.CL↗

arXiv2023

Memristive Memory Enhancement by Device Miniaturization for Neuromorphic Computing

The areal footprint of memristors is a key consideration in material-based neuromorophic computing and large-scale architecture integration. Electronic transport in the most widely investigated memristive devices is mediated by filaments, posing a challenge to their scalability in architecture implementation. Here we present a compelling alternative memristive device and demonstrate that areal downscaling leads to enhancement in memristive memory window, while maintaining analogue behavior, contrary to expectations. Our device designs directly integrated on semiconducting Nb-SrTiO$_3$ allows leveraging electric field effects at edges, increasing the dynamic range in smaller devices. Our findings are substantiated by studying the microscopic nature of switching using scanning transmission electron microscopy, in different resistive states, revealing an interfacial layer whose physical extent is influenced by applied electric fields. The ability of Nb-SrTiO$_3$ memristors to satisfy hardware and software requirements with downscaling, while significantly enhancing memristive functionalities, makes them strong contenders for non-von Neumann computing, beyond CMOS.

cs.ET↗

arXiv2022

On-Shell Action for Type IIB Supergravity and Superstrings on $AdS_5 \times S^5$

AdS/CFT predicts that the value of the on-shell action for type IIB Supergravity (SUGRA) on $AdS_5 \times S^5$ background must be a non-zero number completely determined from the boundary theory. We examine this statement within Sen's formalism for type IIB SUGRA and find that consistency with AdS/CFT requires us to add a specific boundary term to the action. We contrast our resolution with two other resolutions recently proposed in the literature in the context of different approaches to type IIB SUGRA. We explain how our resolution presents a strong benchmark for the possible boundary term of the complete spacetime action for type IIB superstring and how it may possibly lead to a piece of evidence for the strongest form of AdS/CFT conjecture in $AdS_5 \times S^5$. We also comment on the fate of the on-shell action for general self-dual $p$-form fields in Sen's formalism in any curved backgrounds.

hep-th↗

arXiv2020

Irrelevant Deformations of Chiral Bosons

We study $ \text{T}\overline{\text{T}} $ deformations of chiral bosons using the formalism due to Sen. For arbitrary numbers of left- and right-chiral bosons, we find that the $ \text{T}\overline{\text{T}} $-deformed Lagrangian can be computed in closed form, giving rise to a novel non-local action in Sen's formalism. We establish that at the limit of infinite $ \text{T}\overline{\text{T}} $ coupling, the equations of motion of deformed theory exhibits chiral decoupling. We then turn to a discussion of $ \text{T}\overline{\text{T}} $-deformed chiral fermions, and point out that the stress tensor of the $ \text{T}\overline{\text{T}} $-deformed free fermion coincides with the undeformed seed theory. We explain this behaviour of the stress tensor by noting that the deformation term in the action is purely topological in nature and closely resembles the fermionic Wess-Zumino term in the Green-Schwarz formalism. In turn, this observation also explains a puzzle in the literature, viz. why the $ \text{T}\overline{\text{T}} $ deformation of multiple free fermions truncate at linear order. We conclude by discussing the possibility of an interplay between $ \text{T}\overline{\text{T}} $ deformations and bosonisation.

hep-th↗

arXiv2019

Fully Automated Image De-fencing using Conditional Generative Adversarial Networks

Image de-fencing is one of the important aspects of recreational photography in which the objective is to remove the fence texture present in an image and generate an aesthetically pleasing version of the same image without the fence texture. In this paper, we aim to develop an automated and effective technique for fence removal and image reconstruction using conditional Generative Adversarial Networks (cGANs). These networks have been successfully applied in several domains of Computer Vision focusing on image generation and rendering. Our initial approach is based on a two-stage architecture involving two cGANs that generate the fence mask and the inpainted image, respectively. Training of these networks is carried out independently and, during evaluation, the input image is passed through the two generators in succession to obtain the de-fenced image. The results obtained from this approach are satisfactory, but the response time is long since the image has to pass through two sets of convolution layers. To reduce the response time, we propose a second approach involving only a single cGAN architecture that is trained using the ground-truth of fenced de-fenced image pairs along with the edge map of the fenced image produced by the Canny Filter. Incorporation of the edge map helps the network to precisely detect the edges present in the input image, and also imparts it an ability to carry out high quality de-fencing in an efficient manner, even in the presence of a fewer number of layers as compared to the two-stage network. Qualitative and quantitative experimental results reported in the manuscript reveal that the de-fenced images generated by the single-stage de-fencing network have similar visual quality to those produced by the two-stage network. Comparative performance analysis also emphasizes the effectiveness of our approach over state-of-the-art image de-fencing techniques.

cs.CV↗

arXiv2019

Stacking defects in GaP nanowires: Electronic structure and optical properties

Formation of twin boundaries during the growth of semiconductor nanowires is very common. However, the effects of such planar defects on the electronic and optical properties of nanowires are not very well understood. Here, we use a combination of ab initio simulation and experimental techniques to study these effects. Twin boundaries in GaP are shown to act as an atomically-narrow plane of wurtzite phase with a type-I homostructure band alignment. Twin boundaries and stacking faults (wider regions of the wurtzite phase) lead to the introduction of shallow trap states observed in photoluminescence studies. These effects should have a profound impact on the efficiency of nanowire-based devices.

cond-mat.mtrl-sci↗