Searcharxiv⌕ Search

arXiv subjects

Shashank Singh

Publications and source records attributed to Shashank Singh.

At least 19 recordsLinked to original sources
arXiv2026

Benchmarking LLM Judges for Voice-Agent Evaluation: Reliability, Calibration, and Human Oversight

Evaluating conversational voice agents at scale re- quires reliable assessment methods that capture both observ- able interaction quality and the contextual judgment typically provided by human evaluators. We investigate LLM-as-a-Judge evaluation by comparing human judgments with GPT-4.1 and GPT-5 on telecom and retail voice-agent conversations, across conversational quality and safety dimensions. The same interac- tions are scored under three evaluation configurations, p0, p1, and p2, to test whether automated judgments are sensitive to the evaluation setup and whether observed patterns generalize across configurations and judge models. Beyond aggregate agreement, we examine metric-level correlations, evaluator consistency, and systematic human-LLM disagreement to identify which conver- sational attributes can be judged reliably by automation and which remain sensitive to interpretation and context. Effective voice-agent evaluation is also shaped by pipeline-level factors such as speech generation, streaming, and error propagation across ASR, reasoning, and tool-calling stages, motivating our focus on comparing how human and LLM judges score the same interactions end to end. Our results show that LLM- based evaluation can serve as an effective component of large- scale voice-agent assessment, but that its reliability is metric- and configuration-dependent rather than uniform. This pro- vides an empirical framework for identifying which metrics suit automated evaluation and supports hybrid pipelines in which LLM judges handle scalable assessment while human evaluators remain engaged for metrics that demand contextual interpretation and higher-confidence judgment.

cs.AI↗
arXiv2026

Sensitivity Analysis of Dynamic Line Rating for ACSR Conductors using IEEE-738

Dynamic Line Rating (DLR) is a novel technique that enhances the utilization of transmission line capacity. It is nevertheless unclear how much measurement uncertainty in important environmental parameters affects the DLR calculation. Using the IEEE-738 standard, this paper presents a systematic parametric sensitivity analysis of DLR for a 795 kcmil ACSR Drake conductor. The DLR computation encompasses 832 operating points which includes daytime/nighttime solar scenarios, clear/non-clear atmospheric clearness conditions, wind speeds ranging from 0 to 15.25 m/s, and ambient temperatures between 15 and 50 degrees Celsius. According to sensitivity analysis, wind sensitivity tends to decrease at higher ambient temperatures, whereas temperature sensitivity tends to increase with wind speed. Pearson correlation analysis indicates a strong negative linear association between DLR and ambient temperature, and a strong positive linear association with wind speed. A regression model incorporates both individual and interactive effects of wind and temperature, and explains over 93% of the observed DLR variability across all cases. Finally, the observations serve as a guide for operational planning and uncertainty assessment in DLR-based transmission systems.

eess.SY↗
arXiv2025

Bremsstrahlung induced atomic processes

The observed spectra in the collisions of neon (Ne) projectiles of 1.8 and 2.1 MeV with an aluminum target (Al) have been successfully segregated from strong bremsstrahlung backgrounds and then analyzed by comparing the transition energies and rates with the theoretical predictions of the flexible atomic structure code and the general purpose relativistic atomic structure package. The spectra contain Kα, Khα , and Kαα lines. The Kαα emissions are due to two-electron one-photon transitions. Interestingly, the Kαα lines in projectile ions are only seen with 1.8 MeV energy. In contrast, the Kαα lines in the target ions are also well observed with 2.1 MeV energy. Surprisingly, the Al K x-ray line intensities are strongly suppressed, and the Kαα line intensities are unexpectedly enhanced. The underlying physical process is found to be the photoionization caused by intense bremsstrahlung radiation. This photoionization process converts most of the singly ionized K shell states (\approx 82\% at 2.1 MeV) to doubly ionized K shell states. This phenomenon is silently present on many occasions. We take some of such events to validate this remarkable finding. This bremsstrahlung radiation induced secondary ionization process stands as an eye opening incidence to the plasma physics, astronomy and astrophysics communities; may revolutionize these fields of research.

physics.atom-ph↗
arXiv2025

Multiple ionization, fragmentation and dehydrogenation of coronene in collisions with swift protons

Coronene molecules have been bombarded with protons of energy ranging from 100 to 300 keV. The time of flight mass spectra have been recorded using a two stage Wiley McLaren type spectrometer. A significant enhancement in the yields of doubly and triply ionized recoil ions is observed compared to the singly ionized ones. The single, double and triple ionization cross sections are also calculated theoretically using the continuum distorted wave eikonal initial state (CDW EIS) and are compared with the experimental results. The experimental ratios of yields of double to single charged and triple to single charged recoil ions are found to be much higher compared to those for the gaseous atoms. Evaporation peaks corresponding to the loss of several neutral C2H2 molecules are observed for singly, doubly and triply charged coronene recoil ions. Multi fragmentation peaks corresponding to smaller masses of carbohydrates CnHx (n = 3 to 7), appear in the spectra due to higher energy transfer from the projectile to the molecule. The yields of evaporation and fragment products exhibit a pronounced dependence on projectile energy, with a significant decrease observed at higher energies. Dehydrogenetaion i.e. loss of H atoms or H2 molecules are also investigated from the measured spectra. It is observed that hydrogen molecule losses are preferred over H loss in the cation and dication coronene peak structures, with up to three molecules being lost. This observation is in line with some of the predictions and may provide important inputs towards the astrochemistry regarding the observed abundance of H2 in the inter stellar medium.

physics.atom-ph↗
arXiv2023

Spuriosity Didn't Kill the Classifier: Using Invariant Predictions to Harness Spurious Features

To avoid failures on out-of-distribution data, recent works have sought to extract features that have an invariant or stable relationship with the label across domains, discarding "spurious" or unstable features whose relationship with the label changes across domains. However, unstable features often carry complementary information that could boost performance if used correctly in the test domain. In this work, we show how this can be done without test-domain labels. In particular, we prove that pseudo-labels based on stable features provide sufficient guidance for doing so, provided that stable and unstable features are conditionally independent given the label. Based on this theoretical insight, we propose Stable Feature Boosting (SFB), an algorithm for: (i) learning a predictor that separates stable and conditionally-independent unstable features; and (ii) using the stable-feature predictions to adapt the unstable-feature predictions in the test domain. Theoretically, we prove that SFB can learn an asymptotically-optimal predictor without test-domain labels. Empirically, we demonstrate the effectiveness of SFB on real and synthetic data.

cs.LG↗
arXiv2023

Probable Domain Generalization via Quantile Risk Minimization

Domain generalization (DG) seeks predictors which perform well on unseen test distributions by leveraging data drawn from multiple related training distributions or domains. To achieve this, DG is commonly formulated as an average- or worst-case problem over the set of possible domains. However, predictors that perform well on average lack robustness while predictors that perform well in the worst case tend to be overly-conservative. To address this, we propose a new probabilistic framework for DG where the goal is to learn predictors that perform well with high probability. Our key idea is that distribution shifts seen during training should inform us of probable shifts at test time, which we realize by explicitly relating training and test domains as draws from the same underlying meta-distribution. To achieve probable DG, we propose a new optimization problem called Quantile Risk Minimization (QRM). By minimizing the $α$-quantile of predictor's risk distribution over domains, QRM seeks predictors that perform well with probability $α$. To solve QRM in practice, we propose the Empirical QRM (EQRM) algorithm and provide: (i) a generalization bound for EQRM; and (ii) the conditions under which EQRM recovers the causal predictor as $α\to 1$. In our experiments, we introduce a more holistic quantile-focused evaluation protocol for DG and demonstrate that EQRM outperforms state-of-the-art baselines on datasets from WILDS and DomainBed.

stat.ML↗
arXiv2023

Indirect Active Learning

Traditional models of active learning assume a learner can directly manipulate or query a covariate $X$ in order to study its relationship with a response $Y$. However, if $X$ is a feature of a complex system, it may be possible only to indirectly influence $X$ by manipulating a control variable $Z$, a scenario we refer to as Indirect Active Learning. Under a nonparametric model of Indirect Active Learning with a fixed budget, we study minimax convergence rates for estimating the relationship between $X$ and $Y$ locally at a point, obtaining different rates depending on the complexities and noise levels of the relationships between $Z$ and $X$ and between $X$ and $Y$. We also identify minimax rates for passive learning under comparable assumptions. In many cases, our results show that, while there is an asymptotic benefit to active learning, this benefit is fully realized by a simple two-stage learner that runs two passive experiments in sequence.

math.ST↗
arXiv2022

Decoding Attention from Gaze: A Benchmark Dataset and End-to-End Models

Eye-tracking has potential to provide rich behavioral data about human cognition in ecologically valid environments. However, analyzing this rich data is often challenging. Most automated analyses are specific to simplistic artificial visual stimuli with well-separated, static regions of interest, while most analyses in the context of complex visual stimuli, such as most natural scenes, rely on laborious and time-consuming manual annotation. This paper studies using computer vision tools for "attention decoding", the task of assessing the locus of a participant's overt visual attention over time. We provide a publicly available Multiple Object Eye-Tracking (MOET) dataset, consisting of gaze data from participants tracking specific objects, annotated with labels and bounding boxes, in crowded real-world videos, for training and evaluating attention decoding algorithms. We also propose two end-to-end deep learning models for attention decoding and compare these to state-of-the-art heuristic methods.

cs.CV↗
arXiv2022

Optimal Binary Classification Beyond Accuracy

The vast majority of statistical theory on binary classification characterizes performance in terms of accuracy. However, accuracy is known in many cases to poorly reflect the practical consequences of classification error, most famously in imbalanced binary classification, where data are dominated by samples from one of two classes. The first part of this paper derives a novel generalization of the Bayes-optimal classifier from accuracy to any performance metric computed from the confusion matrix. Specifically, this result (a) demonstrates that stochastic classifiers sometimes outperform the best possible deterministic classifier and (b) removes an empirically unverifiable absolute continuity assumption that is poorly understood but pervades existing results. We then demonstrate how to use this generalized Bayes classifier to obtain regret bounds in terms of the error of estimating regression functions under uniform loss. Finally, we use these results to develop some of the first finite-sample statistical guarantees specific to imbalanced binary classification. Specifically, we demonstrate that optimal classification performance depends on properties of class imbalance, such as a novel notion called Uniform Class Imbalance, that have not previously been formalized. We further illustrate these contributions numerically in the case of $k$-nearest neighbor classification

math.ST↗
arXiv2021

Robust Density Estimation under Besov IPM Losses

We study minimax convergence rates of nonparametric density estimation in the Huber contamination model, in which a proportion of the data comes from an unknown outlier distribution. We provide the first results for this problem under a large family of losses, called Besov integral probability metrics (IPMs), that includes $\mathcal{L}^p$, Wasserstein, Kolmogorov-Smirnov, and other common distances between probability distributions. Specifically, under a range of smoothness assumptions on the population and outlier distributions, we show that a re-scaled thresholding wavelet series estimator achieves minimax optimal convergence rates under a wide variety of losses. Finally, based on connections that have recently been shown between nonparametric density estimation under IPM losses and generative adversarial networks (GANs), we show that certain GAN architectures also achieve these minimax rates.

math.ST↗
arXiv2021

K-shell ionization cross sections of Cu, Zn and Ge by 3-5 MeV/U Si-ion bombardment

The K x-ray spectra of different targets (Cu, Zn, and Ge) induced by 3 to 5 MeV/u Si projectile ions have been measured to determine the K-shell ionization cross-section. A significant difference is observed between the measurements and theoretical estimates, where the theoretical ones are about 28-35% of the experimental results. Such difference is reduced to a good extent 51-56% if multiple ionization effects are taken into account. Remaining discrepancy may be attributed to the electron capture contribution.

physics.atom-ph↗
arXiv2021

Continuum-Armed Bandits: A Function Space Perspective

Continuum-armed bandits (a.k.a., black-box or $0^{th}$-order optimization) involves optimizing an unknown objective function given an oracle that evaluates the function at a query point, with the goal of using as few query points as possible. In the most well-studied case, the objective function is assumed to be Lipschitz continuous and minimax rates of simple and cumulative regrets are known in both noiseless and noisy settings. This paper studies continuum-armed bandits under more general smoothness conditions, namely Besov smoothness conditions, on the objective function. In both noiseless and noisy conditions, we derive minimax rates under simple and cumulative regrets. Our results show that minimax rates over objective functions in a Besov space are identical to minimax rates over objective functions in the smallest Hölder space into which the Besov space embeds.

stat.ML↗
arXiv2021

Evaluation of accurate uncertainty of measurement in L subshell ionization cross-section

To have a better understanding of a physical process, a comparison of ex-perimental data with theoretical values is mandatory. The comparison ismeaningful if the uncertainty in the experiment is accounted well. However,it is seldom seen, especially for a complex phenomenon. We take a test casethrough L subshell ionization of atoms by particle impact. Experimentally,x-ray production cross-sections are measured, but ionization cross-sectionsare calculated theoretically. Furthermore, the uncertainty of the x-ray pro-duction cross-section is mainly statistics and detector-efficiency driven. Butionization cross-section involves many other factors because of the relation-ship between the production and ionization cross section, having wide uncer-tainty spectrum. Consequently, determining the measurement uncertainty inL subshell ionization cross-section is always difficult. We have studied thisissue in the simplest way, where the rule of weighted propagation of rela-tive uncertainty is utilised. We notice that larger uncertainties are involvedin atomic parameters relevant to L1(2s1/2) subshell than those associatedwith the other two L2(2p1/2) and L3(2p3/2) subshells. Hence, comparisonbetween theory and experiment would give higher emphasis onL2andL3subshell ionization cross sections. We believe this work aware us that theappropriate uncertainty evaluation is extremely important for providing theright judgment on the data.

physics.atom-ph↗
arXiv2021

Interpretable Sequence Learning for COVID-19 Forecasting

We propose a novel approach that integrates machine learning into compartmental disease modeling to predict the progression of COVID-19. Our model is explainable by design as it explicitly shows how different compartments evolve and it uses interpretable encoders to incorporate covariates and improve performance. Explainability is valuable to ensure that the model's forecasts are credible to epidemiologists and to instill confidence in end-users such as policy makers and healthcare institutions. Our model can be applied at different geographic resolutions, and here we demonstrate it for states and counties in the United States. We show that our model provides more accurate forecasts, in metrics averaged across the entire US, than state-of-the-art alternatives, and that it provides qualitatively meaningful explanatory insights. Lastly, we analyze the performance of our model for different subgroups based on the subgroup distributions within the counties.

cs.LG↗
arXiv2020

Multiclass Classification via Class-Weighted Nearest Neighbors

We study statistical properties of the k-nearest neighbors algorithm for multiclass classification, with a focus on settings where the number of classes may be large and/or classes may be highly imbalanced. In particular, we consider a variant of the k-nearest neighbor classifier with non-uniform class-weightings, for which we derive upper and minimax lower bounds on accuracy, class-weighted risk, and uniform error. Additionally, we show that uniform error bounds lead to bounds on the difference between empirical confusion matrix quantities and their population counterparts across a set of weights. As a result, we may adjust the class weights to optimize classification metrics such as F1 score or Matthew's Correlation Coefficient that are commonly used in practice, particularly in settings with imbalanced classes. We additionally provide a simple example to instantiate our bounds and numerical experiments.

stat.ML↗
arXiv2020

Realization of Self-Demand Response Through Non-Intrusive Load Monitoring Algorithm

The real-time statistics on key consumer parameters and key utility parameters earmark the implementation of demand response (DR) under the smart grid (SG) paradigm. Advanced metering infrastructure (AMI) enables monitoring and control over both key parameters to reflect reliable information on SG. Firstly, this paper aims at a physical realization of an AMI, which comprises a meter data management system (MDMS) supported by smart meters. Secondly, a sliding window based non-intrusive load monitoring algorithm is proposed to illustrate the power consumption pattern. The MDMS features an incremental block--rate tariff structure given by Tamil Nadu state electricity board for the estimation and prediction of electricity bill. Finally, an implementation perspective is presented describing the features of its constituents both theoretically and experimentally, which could envision a consumer-facing grid and exert the proposed self--DR scheme.

eess.SY↗
arXiv2020

FPGA Based SoC Estimator and Constant Current Charging/Discharging Controller for Lead-Acid Battery

The state of charge (SoC) and the rate of charging/discharging current are the vital parameters associated with a battery by which its accurate runtime can be estimated. This paper aims to design a controller which comprises of a field-programmable gate array, back to back connected dc-dc converters, and a resistive touch display based graphical user interface (GUI). The controller estimates SoC and performs constant current charging or discharging of the battery. The implementation of GUI is to input the reference charging/discharging current from the user and to display the SoC. This research delves into battery and obtains its internal parameters by conducting hybrid pulse power characterization test. Furthermore, the obtained internal parameters are processed through an extended Kalman filter which yields the SoC. This controller has large applications in the renewable energy system, battery testing system, smart residential energy management systems, and micro-grids.

eess.SP↗