arXiv subjects

Utkarsh Singh

Publications and source records attributed to Utkarsh Singh.

At least 19 recordsLinked to original sources

arXiv2026

Can Reasoning Models Detect Changes to their Chains of Thought?

There are many reasons one may want to edit a model's chain of thought (CoT) -- e.g., to prefill it with reasoning from a stronger model or to remove steps that may yield unsafe outputs. The success of these interventions plausibly depends on a model's inability to notice them, as the model may alter its behavior if it suspects tampering. In this work, we study whether recent reasoning models are able to detect such interventions on their CoTs under a variety of conditions: both during reasoning and after it, and when prefilled both with their own CoTs and with those of other models. Broadly, we find that (i) models exhibit only very modest detection accuracy; (ii) models struggle to identify *how* their CoT was modified; and (iii) models are about as good at detecting changes to their own CoTs as to those of other models.

cs.AI↗

arXiv2026

Voice of India: A Large-Scale Benchmark for Real-World Speech Recognition in India

Existing Indic ASR benchmarks often use scripted, clean speech and leaderboard driven evaluation that encourages dataset specific overfitting. In addition, strict single reference WER penalizes natural spelling variation in Indian languages, including non standardized spellings of code-mixed English origin words. To address these limitations, we introduce Voice of India, a closed source benchmark built from unscripted telephonic conversations covering 15 major Indian languages across 139 regional clusters. The dataset contains 306230 utterances, totaling 536 hours of speech from 36691 speakers with transcripts accounting for spelling variations. We also analyze performance geographically at the district level, revealing disparities. Finally, we provide detailed analysis across factors such as audio quality, speaking rate, gender, and device type, highlighting where current ASR systems struggle and offering insights for improving real world Indic ASR systems.

cs.CL↗

arXiv2026

Recurrent Quantum Feature Maps for Reservoir Computing

Reservoir computing promises a fast method for handling large amounts of temporal data. This hinges on constructing a good reservoir--a dynamical system capable of transforming inputs into a high-dimensional representation while remembering properties of earlier data. In this work, we introduce a reservoir based on recurrent quantum feature maps where a fixed quantum circuit is reused to encode both current inputs and a classical feedback signal derived from previous outputs. We evaluate the model on the Mackey-Glass time-series prediction task using our recently introduced CP feature map, and find that it achieves lower mean squared error than standard classical baselines, including echo state networks and multilayer perceptrons, while maintaining compact circuit depth and qubit requirements. We further analyze memory capacity and show that the model effectively retains temporal information, consistent with its forecasting accuracy. Finally, we study the impact of realistic noise and find that performance is robust to several noise channels but remains sensitive to two-qubit gate errors, identifying a key limitation for near-term implementations.

quant-ph↗

arXiv2026

Local and Multi-Scale Strategies to Mitigate Exponential Concentration in Quantum Kernels

Fidelity-based quantum kernels provide a direct interface between quantum feature maps and classical kernel methods, but they can exhibit exponential concentration: with increasing system size or circuit expressivity, the Gram matrix approaches the identity and suppresses informative similarity structure. We present an empirical study of two mitigation strategies implemented in Qiskit: (i) local (patch-wise) kernels that aggregate subsystem similarities, and (ii) multi-scale kernels that mix local and global similarity across patch granularities. We benchmark baseline, local, and multi-scale kernels under matched preprocessing, splits, and SVM protocols on several tabular datasets, sweeping the feature dimension $d\in\{4,6,\dots,20\}$. We report concentration diagnostics based on off-diagonal kernel statistics, spectral richness via effective rank, and centered alignment with labels. Across datasets, local and multi-scale constructions consistently mitigate concentration and yield richer kernel spectra relative to the global fidelity baseline, while the impact on classification accuracy depends on the dataset and dimension.

quant-ph↗

arXiv2026

Physics-Integrated Inference for Signal Recovery in Non-Gaussian Regimes

High-performance room-temperature sensing is often limited by non-stationary $1/f$ fluctuations and non-Gaussian stochasticity. In spintronic devices, thermally activated N\'eel switching creates heavy-tailed noise that masks weak signals, defeating linear filters optimized for Gaussian statistics. Here, we introduce a physics-integrated inference framework that decouples signal morphology from stochastic transients using a hierarchical 1D CNN-GRU topology. By learning the temporal signatures of N\'eel relaxation, this architecture reduces the Noise Equivalent Differential Temperature (NEDT) of spintronic Poisson bolometers by a factor of six (233.78 mK to 40.44 mK), effectively elevating room-temperature sensitivity toward cryogenic limits. We demonstrate the framework's universality across the electromagnetic and biological spectrum, achieving a 9-fold error suppression in Radar tracking, a 40\% uncertainty reduction in LiDAR, and a 15.56 dB SNR enhancement in ECG. This hardware-inference coupling recovers deterministic signals from fluctuation-dominated regimes, enabling near-ideal detection limits in noisy edge environments.

physics.optics↗

arXiv2026

Uncooled Poisson Bolometer for High-Speed Event-Based Long-wave Thermal Imaging

Event-based vision provides high-speed, energy-efficient sensing for applications such as autonomous navigation and motion tracking. However, implementing this technology in the long-wave infrared remains a significant challenge. Traditional infrared sensors are hindered by slow thermal response times or the heavy power requirements of cryogenic cooling. Here, we introduce the first event-based infrared detector operating in a Poisson-counting regime. This is realized with a spintronic Poisson bolometer capable of broadband detection from 0.8-14$\mu\text{m}$. In this regime, infrared signals are detected through statistically resolvable changes in stochastic switching events. This approach enables room-temperature operation with high timing resolution. Our device achieves a maximum event rate of 1,250 Hz, surpassing the temporal resolution of conventional uncooled microbolometers by a factor of 4. Power consumption is kept low at 0.2$\mu$W per pixel. This work establishes an operating principle for infrared sensing and demonstrates a pathway toward high-speed, energy-efficient, event-driven thermal imaging.

physics.optics↗

arXiv2026

Ultra-broadband Mid to Long-wave Infrared Spintronic Poisson Bolometer

Infrared detectors have traditionally been divided into two fundamental classes, mid-wave (MWIR, 3-5 um) and long-wave (LWIR, 8-14 um). Integrating MWIR and LWIR within a single device is challenging due to distinct materials, cooling needs, and detection mechanisms, while such integration is critical for improved object recognition, temperature estimation, and environmental sensing. In this work, we demonstrate a Spintronic Poisson (SP) bolometer enabling room-temperature ultra-broadband sensing across 3-14 um. Unlike conventional bolometers that rely on continuous analog signals, the SP bolometer implements a Poisson-counting detection paradigm, encoding temperature in discrete stochastic events, which turns thermal noise from a limitation into the basis of the estimator itself. We fabricate the SP bolometer using a spintronic transduction layer integrated with a plasmonic nanoantenna array to enhance broadband infrared absorption. Using spintronic transduction, the device achieves the noise-equivalent temperature difference (NETD, thermal sensitivity metric) of 80-100 mK at 300 K, surpassing uncooled detectors and approaching cooled technologies. This work establishes a statistical detection paradigm for room-temperature infrared sensing with broad application potential.

physics.optics↗

arXiv2026

Car Drag Coefficient Prediction from 3D Point Clouds Using a Slice-Based Surrogate Model

The automotive industry's pursuit of enhanced fuel economy and performance necessitates efficient aerodynamic design. However, traditional evaluation methods such as computational fluid dynamics (CFD) and wind tunnel testing are resource intensive, hindering rapid iteration in the early design stages. Machine learning-based surrogate models offer a promising alternative, yet many existing approaches suffer from high computational complexity, limited interpretability, or insufficient accuracy for detailed geometric inputs. This paper introduces a novel lightweight surrogate model for the prediction of the aerodynamic drag coefficient (Cd) based on a sequential slice-wise processing of the geometry of the 3D vehicle. Inspired by medical imaging, 3D point clouds of vehicles are decomposed into an ordered sequence of 2D cross-sectional slices along the stream-wise axis. Each slice is encoded by a lightweight PointNet2D module, and the sequence of slice embeddings is processed by a bidirectional LSTM to capture longitudinal geometric evolution. The model, trained and evaluated on the DrivAerNet++ dataset, achieves a high coefficient of determination (R^2 > 0.9528) and a low mean absolute error (MAE approx 6.046 x 10^{-3}) in Cd prediction. With an inference time of approximately 0.025 seconds per sample on a consumer-grade GPU, our approach provides fast, accurate, and interpretable aerodynamic feedback, facilitating more agile and informed automotive design exploration.

cs.CV↗

arXiv2025

Dynamic Screening Effects on Auger Recombination in Metal-Halide Perovskites

The performance of modern light-emitting technologies, from lasers to LEDs, is limited by nonradiative losses, with Auger recombination being the dominant channel at device-relevant carrier densities. Reliable modeling of this process is essential, yet conventional treatments neglect dynamic dielectric effects, limiting the predictive reliability at operating conditions. We develop a general framework that incorporates the frequency-dependent screened Coulomb interaction $W_{00}(\mathbf{q},\omega)$, computed from low-scaling \textit{GW}, into both direct and phonon-assisted Auger amplitudes. Demonstrated on orthorhombic $\gamma$-CsPbI$_3$ (band gap $E_g\approx1.73$ eV) and $\gamma$-CsSnI$_3$ ($E_g\approx1.30$ eV), the approach shows that dynamic screening enhances the dielectric response, lowering the room-temperature Auger coefficient by $\sim$50-60 %. This renormalization shifts the crossover between radiative and nonradiative recombination by nearly a factor of two in carrier density. Dynamic dielectric screening thus emerges as a quantitative determinant of Auger recombination, offering a transferable framework for predictive modeling across polar semiconductors where frequency-independent screening models are inadequate.

cond-mat.mtrl-sci↗

arXiv2025

Long wave infrared detection using probabilistic spintronic bolometer arrays

The use of probabilistic spintronic devices for infrared radiation detection has introduced a shift in approach to thermal imaging. The integration of probabilistic magnetic tunnel junctions with infrared plasmonic nano-antennas achieves high-sensitivity digital-mode infrared sensors at room temperature. Here, we present a scalable approach towards multipixel plasmonic-spintronic bolometer array fabrication and readout. We fabricate proof-of-concept 2x2 row-column multiplexed probabilistic plasmonic sprintronic arrays and show their response to long-wave infrared radiation (8-14um) with high readout speeds (10K-1M counts per second). These spintronic, ultrafast, nanoscale (SUN) bolometers can result in novel high-pixel density CMOS compatible infrared detection platforms. Our work provides a broadband (9kHz to 3GHz) readout platform for future digital probabilistic detector applications. Furthermore, our approach addresses a key challenge associated with scaling infrared pixel sizes that can drive progress towards high pixel density detector arrays for infrared sensing and microscopy applications.

physics.ins-det↗

arXiv2025

Lightweight Transformers for Zero-Shot and Fine-Tuned Text-to-SQL Generation Using Spider

Text-to-SQL translation enables non-expert users to query relational databases using natural language, with applications in education and business intelligence. This study evaluates three lightweight transformer models - T5-Small, BART-Small, and GPT-2 - on the Spider dataset, focusing on low-resource settings. We developed a reusable, model-agnostic pipeline that tailors schema formatting to each model's architecture, training them across 1000 to 5000 iterations and evaluating on 1000 test samples using Logical Form Accuracy (LFAcc), BLEU, and Exact Match (EM) metrics. Fine-tuned T5-Small achieves the highest LFAcc (27.8%), outperforming BART-Small (23.98%) and GPT-2 (20.1%), highlighting encoder-decoder models' superiority in schema-aware SQL generation. Despite resource constraints limiting performance, our pipeline's modularity supports future enhancements, such as advanced schema linking or alternative base models. This work underscores the potential of compact transformers for accessible text-to-SQL solutions in resource-scarce environments.

cs.CL↗

arXiv2025

A Resource Efficient Quantum Kernel

Quantum processors may enhance machine learning by mapping high-dimensional data onto quantum systems for processing. Conventional feature maps, for encoding data onto a quantum circuit are currently impractical, as the number of entangling gates scales quadratically with the dimension of the dataset and the number of qubits. In this work, we introduce a quantum feature map designed to handle high-dimensional data with a significantly reduced number of qubits and entangling operations. Our approach preserves essential data characteristics while promoting computational efficiency, as evidenced by extensive experiments on benchmark datasets that demonstrate a marked improvement in both accuracy and resource utilization when using our feature map as a kernel for characterization, as compared to state-of-the-art quantum feature maps. Our noisy simulation results, combined with lower resource requirements, highlight our map's ability to function within the constraints of noisy intermediate-scale quantum devices. Through numerical simulations and small-scale implementation on a superconducting circuit quantum computing platform, we demonstrate that our scheme performs on par or better than a set of classical algorithms for classification. While quantum kernels are typically stymied by exponential concentration, our approach is affected with a slower rate with respect to both the number of qubits and features, which allows practical applications to remain within reach. Our findings herald a promising avenue for the practical implementation of quantum machine learning algorithms on near future quantum computing platforms.

quant-ph↗

arXiv2024

Coherent Feed Forward Quantum Neural Network

Quantum machine learning, focusing on quantum neural networks (QNNs), remains a vastly uncharted field of study. Current QNN models primarily employ variational circuits on an ansatz or a quantum feature map, often requiring multiple entanglement layers. This methodology not only increases the computational cost of the circuit beyond what is practical on near-term quantum devices but also misleadingly labels these models as neural networks, given their divergence from the structure of a typical feed-forward neural network (FFNN). Moreover, the circuit depth and qubit needs of these models scale poorly with the number of data features, resulting in an efficiency challenge for real-world machine-learning tasks. We introduce a bona fide QNN model, which seamlessly aligns with the versatility of a traditional FFNN in terms of its adaptable intermediate layers and nodes, absent from intermediate measurements such that our entire model is coherent. This model stands out with its reduced circuit depth and number of requisite C-NOT gates to outperform prevailing QNN models. Furthermore, the qubit count in our model remains unaffected by the data's feature quantity. We test our proposed model on various benchmarking datasets such as the diagnostic breast cancer (Wisconsin) and credit card fraud detection datasets. We compare the outcomes of our model with the existing QNN methods to showcase the advantageous efficacy of our approach, even with a reduced requirement on quantum resources. Our model paves the way for application of quantum neural networks to real relevant machine learning problems.

quant-ph↗

arXiv2023

memorAIs: an Optical Character Recognition and Rule-Based Medication Intake Reminder-Generating Solution

Memory-based medication non-adherence is an unsolved problem that is responsible for considerable disease burden in the United States. Digital medication intake reminder solutions with minimal onboarding requirements that are usable at the point of medication acquisition may help to alleviate this problem by offering a low barrier way to help people remember to take their medications. In this paper, we propose memorAIs, a digital medication intake reminder solution that mitigates onboarding friction by leveraging optical character recognition strategies for text extraction from medication bottles and rule based expressions for text processing to create configured medication reminders as local device calendar invitations. We describe our ideation and development process, as well as limitations of the current implementation. memorAIs was the winner of the Patient Safety award at the 2023 Columbia University DivHacks Hackathon, presented by the Patient Safety Technology Challenge, sponsored by the Pittsburgh Regional Health Initiative.

cs.IR↗

arXiv2023

Exploring Magnetism of Lead-free Halide Double Perovskites: A High-Throughput First-Principles Study

We have performed a comprehensive, first-principles high-throughput study of the magnetic properties of halide double perovskites, $Cs_2BB^\prime Cl_6$, with magnetic ions occupying one or both B and B$^\prime$ sites. Our findings indicate a general tendency for these materials to exhibit antiferromagnetic ordering with low N\'eel temperatures. At the same time, we reveal a few potential candidates that predicted to be ferromagnetic with relatively high Curie temperatures. Achieving ferromagnetic coupling might be feasible via simultaneously alloying at B and B$^\prime$ sites with magnetic 3d and non-magnetic 5d ions. With this approach, we discover that $Cs_2HgCrCl_6$, $Cs_2AgNiCl_6$ and $Cs_2AuNiCl_6$ have high Curie temperatures relative to their peers, with the latter two exhibiting half metallic behaviour. Further, this study illuminates the underpinning mechanism of magnetic exchange interactions in halide double perovskites, enabling a deeper understanding of their magnetic behaviour. Our findings, especially the discovery of the compounds with robust half-metallic properties and high Curie temperatures holds promise for potential applications in the field of spintronics.

cond-mat.mtrl-sci↗

arXiv2023

The role of hosting capacity study in power system advancements: a review

The fast depletion of conventional energy sources due to increased energy demands and environmental concern has motivated power utilities to integrate more renewable energy sources into their power systems. Due to the intermittent nature and low or non-existent inertial response of these sources, a high penetration of RESs can lead to various issues in the operation of power systems such as oscillations in power system voltage and frequency, increased harmonic distortion, failure of protective equipment, overloading of transformers and feeders, and increased line losses. Such problems arise when the hosting capacity, defined as the maximum RES capacity that can be installed without having any technical and operational problems, of the network exceeds its limit due to the increased integration of RESs to the existing network. This paper reviews the progress made in HC assessment and enhancement of electrical networks research and development since its inception. Attempts are also made to highlight the current and future issues involved in HC technology for the development of an affordable, in-exhaustive, clean and reliable power supply for longer term benefits

eess.SY↗

arXiv2022

Electronic structure of the magnetic halide double perovskites Cs2(Ag,Na)FeCl6 from first-principles

A family of magnetic halide double perovskites (HDPs) have recently attracted attention due to their potential to broaden application areas of halide double perovskites into e.g. spintronics. Up to date the theoretical modelling of these systems have relied on primitive approximations to the density functional theory (DFT). In this paper, we study structural, electronic and magnetic properties of the Fe$^{3+}$-containing HDPs Cs$_2$AgFeCl$_6$ and Cs$_2$NaFeCl$_6$ using a combination of more advanced DFT-based methods, including DFT+U, hybrid-DFT and treatments of various magnetic states. We examine the effect of varying the effective Hubbard parameter, U$_{eff}$, in DFT+U and the mixing-parameter, $\alpha$, in hybrid DFT on the electronic structure and structural properties. Our results reveal a set of localized Fe(d) states that are highly sensitive to these parameters. Cs$_2$AgFeCl$_6$ and Cs$_2$NaFeCl$_6$ are both antiferromagnets with Ne\'el temperatures well below room temperature and are thus in their paramagnetic (PM) state at the external conditions relevant to most applications. Therefore, we have examined the effect of disordered magnetism on the electronic structure of these systems and find that while Cs$_2$NaFeCl$_6$ is largely unaffected, Cs$_2$AgFeCl$_6$ shows significant renormalization of its electronic band structure.

cond-mat.mtrl-sci↗

arXiv2020

Monitoring Large Crowds With WiFi: A Privacy-Preserving Approach

This paper presents a crowd monitoring system based on the passive detection of probe requests. The system meets strict privacy requirements and is suited to monitoring events or buildings with a least a few hundreds of attendees. We present our counting process and an associated mathematical model. From this model, we derive a concentration inequality that highlights the accuracy of our crowd count estimator. Then, we describe our system. We present and discuss our sensor hardware, our computing system architecture, and an efficient implementation of our counting algorithm -- as well as its space and time complexity. We also show how our system ensures the privacy of people in the monitored area. Finally, we validate our system using nine weeks of data from a public library endowed with a camera-based counting system, which generates counts against which we compare those of our counting system. This comparison empirically quantifies the accuracy of our counting system, thereby showing it to be suitable for monitoring public areas. Similarly, the concentration inequality provides a theoretical validation of the system.

eess.SY↗