arXiv Science⌕ Search

arXiv subjects

Santosh Kumar

Publications and source records attributed to Santosh Kumar.

At least 37 records · Page 2Linked to original sources

How Many Times Do People Usually Experience Different Kinds of Stressors Each Day?

Understanding how frequently people experience different kinds of daily stressors is crucial for interpreting stress exposure and informing mental health care. But it can't be directly estimated from current assessment methods, such as diaries, end-of-day interviews, and ecological momentary assessments (EMA), that use sparse sampling to limit participant burden, and a structured response format for uniformity. In this paper, we utilize stressor data collected in a 100-day field study with 68 participants that adopted wearable-triggered prompts and a freeform format to solicit stressors soon after they occurred, but limited its prompts to a small subset to keep the burden low. We develop asymptotic models to estimate the latent frequency of different kinds of real-life stressors that address sample sparsity and sampling bias. We find that people experience 5.39 stressors per day, on average. The top three are related to work (1.76/day), health (0.59/day), and transportation (0.55/day). These estimates offer a principled benchmark for interpreting individual stressor loads. They can also inform mental health care treatments and interventions by establishing population-level baselines.

cs.HC↗

Pulse-PPG: An Open-Source Field-Trained PPG Foundation Model for Wearable Applications Across Lab and Field Settings

Photoplethysmography (PPG)-based foundation models are gaining traction due to the widespread use of PPG in biosignal monitoring and their potential to generalize across diverse health applications. In this paper, we introduce Pulse-PPG, the first open-source PPG foundation model trained exclusively on raw PPG data collected over a 100-day field study with 120 participants. Existing PPG foundation models are either open-source but trained on clinical data or closed-source, limiting their applicability in real-world settings. We evaluate Pulse-PPG across multiple datasets and downstream tasks, comparing its performance against a state-of-the-art foundation model trained on clinical data. Our results demonstrate that Pulse-PPG, trained on uncurated field data, exhibits superior generalization across clinical and mobile health applications in both lab and field settings. This suggests that exposure to real-world variability enables the model to learn fine-grained representations, making it more adaptable across tasks. Furthermore, pre-training on field data surprisingly outperforms its pre-training on clinical data in many tasks, reinforcing the importance of training on real-world, diverse datasets. To encourage further advancements in robust foundation models leveraging field data, we plan to release Pulse-PPG, providing researchers with a powerful resource for developing more generalizable PPG-based models.

cs.LG↗

Dense Associative Memory in a Nonlinear Optical Hopfield Neural Network

Modern Hopfield Neural Networks (HNNs), also known as Dense Associative Memories (DAMs), enhance the performance of simple recurrent neural networks by leveraging the nonlinearities in their energy functions. They have broad applications in combinatorial optimization, high-capacity memory storage, deep learning transformers, and correlated pattern recognition. Thus far, research on DAMs has been primarily theoretical, with implementations limited to CPUs and GPUs. In this work, for the first time to our knowledge, we propose and experimentally demonstrate a nonlinear optical Hopfield neural network (NOHNN) system for realizing DAMs using correlated patterns. Our NOHNN incorporates effective 2-body and 4-body interactions in its energy function. The inclusion of 4-body interaction scores a minimum ten-fold improvement in the number of uncorrelated patterns that can be stored and retrieved, significantly surpassing the traditional capacity limit for traditional HNNs. For correlated patterns, depending on their average correlation, up to 50 times more patterns can be stored compared to traditional HNNs. To test the system's robustness, the benchmark testing is performed on MNIST handwritten digit patterns. The results show a 5.5 times improvement in the pattern storage along with the retrieval of cleaner and less noisy patterns. These results highlight the potential of nonlinear optical DAMs for practical applications in challenging big-data optimization, computer vision and graph network tasks.

physics.optics↗

Multi Moire Networks in Engineered Lateral Hetero-Bilayers: Programmable Phononic Reconfiguration and Second Harmonic Generation

Moire engineering in two-dimensional transition metal dichalcogenides enables access to correlated quantum phenomena. Realizing such effects demands simultaneous control over twist angle and material composition to modulate phonons, excitons, and their interactions. However, most studies rely on exfoliated flakes, limiting scalability and systematic exploration. Here, we demonstrate a scalable multi-moire network by vertically stacking CVD-grown monolayer lateral heterostructures. Signatures of moire non-rigidity, including phonon frequency softening, linewidth broadening, and strain localization, are attributed to two lattice relaxation modes; rotational reconstruction and volumetric dilation. Micro-angle-resolved photoemission spectroscopy reveals that interfacial orbital interactions modulate interlayer coupling. At aligned angles, molybdenum diselenides exhibit reduced valley polarization and Davydov splitting, indicating strain-induced symmetry breaking and chiral phonon effects. Notably, SHG modulation was obderved with variation in twist angle due to lower coherence and band-offset-driven phase delay. First-principles calculations support these findings. This work provides a route to programmable, scalable multi-moire platforms for opto-straintronics, quantum sensing, and on-chip photonics.

cond-mat.mes-hall↗

Spectral statistics of interpolating random circulant matrix and its applications to random circulant graphs

We consider a versatile matrix model of the form ${\bf A}+i {\bf B}$, where ${\bf A}$ and ${\bf B}$ are real random circulant matrices with independent but, in general, nonidentically distributed Gaussian entries. For this model, we derive exact results for the joint probability density function and find that it is a multivariate Gaussian. Arbitrary order marginal density therefore also readily follows. It is demonstrated that by adjusting the averages and variances of the Gaussian elements of ${\bf A}$ and ${\bf B}$, we can interpolate between a remarkably wide range of eigenvalue distributions in the complex plane. In particular, we can examine the crossover between a random real circulant matrix and a random complex circulant matrix. We also extend our study to include Wigner-like and Wishart-like matrices constructed from our general random circulant matrix. To validate our analytical findings, Monte Carlo simulations are conducted, which confirm the accuracy of our results. Additionally, we compare our analytical results with the spectra of adjacency matrices from various random circulant graphs. Despite the difference in entry distributions-Gaussian in our model and non-Gaussian in the adjacency matrices-the densities show excellent agreement in the large-dimension limit.

math-ph↗

Layered semiconductors integrated with polyimide thin films for high-quality valleytronic and quantum-photonic systems

Dielectric integration of layered semiconductors is a prerequisite for fabricating high-quality optoelectronic, valleytronic, and quantum-photonic devices. While hexagonal boron nitride (hBN) is the current benchmark dielectric, exploration of the most suitable dielectric materials covering the complete substrates continues to expand. This work demonstrates the formation of high optical-quality excitons in two widely explored layered semiconductors, WSe$_2$ and WS$_2$, integrated into polyimide (PI) thin films of thicknesses $\approx$500 nm. The photoluminescence (PL) studies at $T$ = 296 K show the formation of neutral excitons $\left(X^0\right)$ and trions in fully-PI-encapsulated 1L-WSe$_2$ with 2-sigma ($2σ$) spatial-inhomogeneity of 4.5 (3.4) meV in $X^0$ emission energy (linewidth), which is $\approx$1/3rd (1/5th), respectively, that of inhomogeneity measured in fully-hBN-encapsulated 1L-WSe$_2$. A smaller $2σ$ of 2.1 (2.3) meV in $X^0$ emission energy (linewidth) has been shown for fully-PI-encapsulated 1L-WS$_2$. Polarization-resolved and excitation power-dependent PL measurements of PI-isolated 1L-TMDs at $T$ = 4 K further reveal formations of high-quality neutral-biexcitons and negatively-charged biexcitons, with degrees of valley-polarization up to 21$\%$ under non-resonant excitation. Furthermore, the fully-PI-encapsulated 1L-WSe$_2$ also hosts single quantum emitters with narrow linewidths and high-spectral stability. This work indicates that PI thin films may serve the purpose of high-quality dielectric material for integrating the layered materials on a wafer scale.

cond-mat.mes-hall↗

Superheavy Nuclei and the Changing Face of Nuclear Magicity

Using a relativistic mean field formalism, we analyzed the magic number sequence for finite nuclei in the superheavy valley. The result for the IOPB-I parameter set is compared with the well-known NL3 force. The magic numbers obtained from IOPB-I and NL3 interactions are found to be similar. Analysing the single-particle levels and the number of nucleons occupied in it, we find the close shell sequence as 2, 8, 18, 34, 50, 58, 80, 82, 92, 114, 120, 120, 138, 164, 172, 184 and 198 for the $^{318}{120}$ mass region. Again, with a careful inspection, we noticed large shell gaps at nucleon numbers 2, 8, 18, 34, 50, 58, 80, 92, 120, 138, 164, 172, 184, and 198, which may be considered as the magic number sequence for the superheavy nuclei. This change may be due to the shape change of the nuclear potential as compared to the stability valley.

nucl-th↗

Wearable Meets LLM for Stress Management: A Duoethnographic Study Integrating Wearable-Triggered Stressors and LLM Chatbots for Personalized Interventions

We use a duoethnographic approach to study how wearable-integrated LLM chatbots can assist with personalized stress management, addressing the growing need for immediacy and tailored interventions. Two researchers interacted with custom chatbots over 22 days, responding to wearable-detected physiological prompts, recording stressor phrases, and using them to seek tailored interventions from their LLM-powered chatbots. They recorded their experiences in autoethnographic diaries and analyzed them during weekly discussions, focusing on the relevance, clarity, and impact of chatbot-generated interventions. Results showed that even though most events triggered by the wearable were meaningful, only one in five warranted an intervention. It also showed that interventions tailored with brief event descriptions were more effective than generic ones. By examining the intersection of wearables and LLM, this research contributes to developing more effective, user-centric mental health tools for real-time stress relief and behavior change.

cs.HC↗

Large trion binding energy in monolayer WS$_2$ via strain-enhanced electron-phonon coupling

Transition metal dichalcogenides and related layered materials in their monolayer and a few layers thicknesses regime provide a promising optoelectronic platform for exploring the excitonic- and many-body physics. Strain engineering has emerged as a potent technique for tuning the excitonic properties favorable for exciton-based devices. We have investigated the effects of nanoparticle-induced local strain on the optical properties of exciton, $X^0$, and trion, $X^\text{-}$, in monolayer WS$_2$. Biaxial tensile strain up to 2.0% was quantified and verified by monitoring the changes in three prominent Raman modes of WS$_2$: E${^1_{2g}}$($Γ$), A$_{1g}$, and 2LA(M). We obtained a remarkable increase of 34 meV in $X^\text{-}$ binding energy with an average tuning rate of 17.5 $\pm$ 2.5 meV/% strain across all the samples irrespective of the surrounding dielectric environment of monolayer WS$_2$ and the sample preparation conditions. At the highest tensile strain of $\approx$2%, we have achieved the largest binding energy $\approx$100 meV for $X^\text{-}$, leading to its enhanced emission intensity and thermal stability. By investigating strain-induced linewidth broadening and deformation potentials of both $X^0$ and $X^\text{-}$ emission, we elucidate that the increase in $X^\text{-}$ binding energy is due to strain-enhanced electron-phonon coupling. This work holds relevance for future $X^\text{-}$-based nano-opto-electro-mechanical systems and devices.

cond-mat.mes-hall↗

Machine Learning Algorithms for Detecting Mental Stress in College Students

In today's world, stress is a big problem that affects people's health and happiness. More and more people are feeling stressed out, which can lead to lots of health issues like breathing problems, feeling overwhelmed, heart attack, diabetes, etc. This work endeavors to forecast stress and non-stress occurrences among college students by applying various machine learning algorithms: Decision Trees, Random Forest, Support Vector Machines, AdaBoost, Naive Bayes, Logistic Regression, and K-nearest Neighbors. The primary objective of this work is to leverage a research study to predict and mitigate stress and non-stress based on the collected questionnaire dataset. We conducted a workshop with the primary goal of studying the stress levels found among the students. This workshop was attended by Approximately 843 students aged between 18 to 21 years old. A questionnaire was given to the students validated under the guidance of the experts from the All India Institute of Medical Sciences (AIIMS) Raipur, Chhattisgarh, India, on which our dataset is based. The survey consists of 28 questions, aiming to comprehensively understand the multidimensional aspects of stress, including emotional well-being, physical health, academic performance, relationships, and leisure. This work finds that Support Vector Machines have a maximum accuracy for Stress, reaching 95\%. The study contributes to a deeper understanding of stress determinants. It aims to improve college student's overall quality of life and academic success, addressing the multifaceted nature of stress.

cs.LG↗

Measurements of absolute bandgap deformation-potentials of optically-bright bilayer WSe$_2$

Bilayers of transition-metal dichalcogenides show many exciting features, including long-lived interlayer excitons and wide bandgap tunability using strain. Not many investigations on experimental determinations of deformation potentials relating changes in optoelectronic properties of bilayer WSe$_2$ with the strain are present in the literature. Our experimental study focuses on three widely investigated high-symmetry points, K$_{c}$, K$_{v}$, and Q$_{c}$, where subscript c (v) refers to the conduction (valence) band, in the Brillouin zone of bilayer WSe$_2$. Using local biaxial strains produced by nanoparticle stressors, a theoretical model, and by performing the spatially- and spectrally-resolved photoluminescence measurements, we determine absolute deformation potential of -5.10 $\pm$ 0.24 eV for Q$_{c}$-K$_{v}$ indirect bandgap and -8.50 $\pm$ 0.92 eV for K$_{c}$-K$_{v}$ direct bandgap of bilayer WSe$_2$. We also show that $\approx$0.9% biaxial tensile strain is required to convert an indirect bandgap bilayer WSe$_2$ into a direct bandgap semiconductor. Moreover, we also show that a relatively small amount of localized strain $\approx$0.4% is required to make a bilayer WSe$_2$ as optically bright as an unstrained monolayer WSe$_2$. The bandgap deformation potentials measured here will drive advances in flexible electronics, sensors, and optoelectronic- and quantum photonic- devices through precise strain engineering.

cond-mat.mes-hall↗

Computing marginal eigenvalue distributions for the Gaussian and Laguerre orthogonal ensembles

The Gaussian and Laguerre orthogonal ensembles are fundamental to random matrix theory, and the marginal eigenvalue distributions are basic observable quantities. Notwithstanding a long history, a formulation providing high precision numerical evaluations for $N$ large enough to probe asymptotic regimes, has not been provided. An exception is for the largest eigenvalue, where there is a formalism due to Chiani which uses a combination of the Pfaffian structure underlying the ensembles, and a recursive computation of the matrix elements. We augment this strategy by introducing a generating function for the conditioned gap probabilities. A finite Fourier series approach is then used to extract the sequence of marginal eigenvalue distributions as a linear combination of Pfaffians, with the latter then evaluated using an efficient numerical procedure available in the literature. Applications are given to illustrating various asymptotic formulas, local central limit theorems, and central limit theorems, as well as to probing finite size corrections. Further, our data indicates that the mean values of the marginal distributions interlace with the zeros of the Hermite polynomial (Gaussian ensemble) and a Laguerre polynomial (Laguerre ensemble).

math-ph↗

Chip-integrated Spectroscopy Capable of Temperature Retrieval

We demonstrate a chip-integrated emission spectroscope capable of retrieving the temperature of the light sources. It consists of a single photon detector with low dark counts and a sweeping on-chip filter with 2 pm spectral resolution in the visible and near-infrared regimes. With wildfire sensing applications in mind, we test our system with a hollow cathode lamp to simulate the K-line emission, and show how the models of Doppler and collision broadening in the plasma can be used for temperature retrieval. With favorable device parameters, high spectral resolution, and a novel temperature retrieval capability, our technique may find broad applications in environmental monitoring, astrophysics, plasma physics, and so on.

physics.optics↗

A Robust Super-Resolution Classifier by Nonlinear Optics

Spatial-mode projective measurements could achieve super-resolution in remote sensing and imaging, yet their performance is usually sensitive to the parameters of the target scenes. We propose and demonstrate a robust classifier of close-by light sources by using optimized mode projection via nonlinear optics. Contrary to linear-optics based methods using the first few Hermite-Gaussian modes for the projection, here the projection modes are optimally tailored by shaping the pump wave to drive the nonlinear optical process. This minimizes modulation losses and allows high flexibility in designing those modes for robust and efficient measurements. We test this classifier on discriminating one light source and two sources separated well within the Rayleigh limit without prior knowledge of the exact centroid or brightness. Our results show a classification fidelity of over 80% even when the centroid is misaligned by half the source separation, or when one source is four times stronger than the other.

physics.optics↗

First Photon Machine Learning

Quantum techniques are expected to revolutionize how information is acquired, exchanged, and processed. Yet it has been a challenge to realize and measure their values in practical settings. We present first photon machine learning as a new paradigm of neural networks and establish the first unambiguous advantage of quantum effects for artificial intelligence. By extending the physics behind the double-slit experiment for quantum particles to a many-slit version, our experiment finds that a single photon can perform image recognition at around $30\%$ fidelity, which beats by a large margin the theoretical limit of what a similar classical system can possibly achieve (about 24\%). In this experiment, the entire neural network is implemented in sub-attojoule optics and the equivalent per-calculation energy cost is below $10^{-24}$ joule, highlighting the prospects of quantum optical machine learning for unparalleled advantages in speed, capacity, and energy efficiency.

quant-ph↗

Exact mean and variance of the squared Hellinger distance for random density matrices

The Hellinger distance between quantum states is a significant measure in quantum information theory, known for its Riemannian and monotonic properties. It is also easier to compute than the Bures distance, another measure that shares these properties. In this work, we derive the mean and variance of the Hellinger distance between pairs of density matrices, where one or both matrices are random. Along the way, we also obtain exact results for the mean affinity and mean square affinity. The first two cumulants of the Hellinger distance allow us to propose an approximation for the corresponding probability density function based on the gamma distribution. Our analytical results are corroborated through Monte Carlo simulations, showing excellent agreement.

quant-ph↗

Beyond Nearest-neighbour Universality of Spectral Fluctuations in Quantum Chaotic and Complex Many-body Systems

Discerning chaos in quantum systems is an important problem as the usual route of Lyapunov exponents in classical systems is not straightforward in quantum systems. A standard route is the comparison of statistics derived from model physical systems to those from random matrix theory (RMT) ensembles, of which the most popular is the nearest-neighbour-spacings distribution (NNSD), which almost always shows good agreement with chaotic quantum systems. However, even in these cases, the long-range statistics (like number variance, spectral rigidity etc.), which are also more difficult to calculate, often show disagreements with RMT. As such, a more stringent test for chaos in quantum systems, via an analysis of intermediate-range statistics is needed, which will additionally assess the extent of agreement with RMT universality. In this paper, we deduce the effective level-repulsion parameters and the corresponding Wigner-surmise-like results of the next-nearest-neighbor spacing distribution (nNNSD) for integrable systems (semi-Poissonian statistics) as well as the three classical quantum-chaotic Wigner-Dyson regimes, by stringent comparisons to numerical RMT models and benchmarking against our exact analytical results for $3\times 3$ Gaussian matrix models, along with a semi-analytical form for the nNNSD in the Orthogonal-to-Unitary symmetry crossover. To illustrate the robustness of these RMT based results, we test these predictions against the nNNSD obtained from quantum chaotic models as well as disordered lattice spin models. This reinforces the Bohigas-Giannoni-Schmit and the Berry-Tabor conjectures, extending the associated universality to longer range statistics. In passing, we also highlight the equivalence of nNNSD in the apparently distinct Orthogonal-to-Unitary and diluted-Symplectic-to-Unitary crossovers.

cond-mat.stat-mech↗

Temporally Multi-Scale Sparse Self-Attention for Physical Activity Data Imputation

Wearable sensors enable health researchers to continuously collect data pertaining to the physiological state of individuals in real-world settings. However, such data can be subject to extensive missingness due to a complex combination of factors. In this work, we study the problem of imputation of missing step count data, one of the most ubiquitous forms of wearable sensor data. We construct a novel and large scale data set consisting of a training set with over 3 million hourly step count observations and a test set with over 2.5 million hourly step count observations. We propose a domain knowledge-informed sparse self-attention model for this task that captures the temporal multi-scale nature of step-count data. We assess the performance of the model relative to baselines and conduct ablation studies to verify our specific model designs.

cs.LG↗