arXiv subjects

Rahul Sharma

Publications and source records attributed to Rahul Sharma.

At least 55 records · Page 3Linked to original sources

arXiv2025

Intriguing nature of AM Her type candidate CXOU J204734.8+300105

The detection and characterization of periodic X-ray signals are crucial for identifying new compact objects and studying the mechanisms powering their emission. We report on the timing and spectral variability of CXOU J204734.8+300105, a proposed eclipsing polar-type cataclysmic variable (CV) candidate. This source has been observed once with Chandra and twice with XMM-Newton, revealing several intriguing and conflicting features in its X-ray emission. The Chandra observation showed a periodicity of $\sim$6000 s with an eclipse-like feature. The X-ray light curve from 2017 XMM-Newton observation showed a period of $\sim$2000 seconds without any apparent eclipse, while the simultaneous optical light curve from OM showed a period of $\sim$6000 seconds. This variability raises questions about the true nature of the source. Spectral analysis indicates a multi-component emission and emission lines due to Fe. The spectral characteristics are consistent with those observed in other CV systems. Additionally, we identified optical and near-infrared counterparts from various catalogues. Our findings suggest a dynamic and evolving accretion environment of CXOU J204734.8+300105.

astro-ph.HE↗

arXiv2025

DeduCE: Deductive Consistency as a Framework to Evaluate LLM Reasoning

Despite great performance on Olympiad-level reasoning problems, frontier large language models can still struggle on high school math when presented with novel problems outside standard benchmarks. Going beyond final accuracy, we propose a deductive consistency metric to analyze chain-of-thought output from language models (LMs).Formally, deductive reasoning involves two subtasks: understanding a set of input premises and inferring the conclusions that follow from them. The proposed metric studies LMs' performance on these subtasks, with the goal of explaining LMs' reasoning errors on novel problems: how well do LMs understand input premises with increasing context lengths, and how well can they infer conclusions over multiple reasoning hops? Since existing benchmarks may be memorized, we develop a pipeline to evaluate LMs' deductive consistency on novel, perturbed versions of benchmark problems. On novel grade school math problems (GSM-8k), we find that LMs are fairly robust to increasing number of input premises, but suffer significant accuracy decay as the number of reasoning hops is increased. Interestingly, these errors are masked in the original benchmark as all models achieve near 100% accuracy. As we increase the number of solution steps using a synthetic dataset, prediction over multiple hops still remains the major source of error compared to understanding input premises. Other factors, such as shifts in language style or natural propagation of early errors do not explain the trends. Our analysis provides a new view to characterize LM reasoning -- as computations over a window of input premises and reasoning hops -- that can provide unified evaluation across problem domains.

cs.CL↗

arXiv2025

Sidebands to mHz QPOs in 4U 1626$-$67 in the second spin-down state

We report results from an $AstroSat$ Target-of-Opportunity (ToO) observation of 4U 1626$-$67, performed on 2023 May 18, soon after the discovery of torque reversal to spin-down in the source. The X-ray emission exhibited significant dependence on both energy and torque state. This work highlights the comparison of timing features of 4U 1626$-$67 with a previous $AstroSat$ observation from 2018, when the neutron star was in the spin-up state. The power density spectrum (PDS) of the 2023 observation comprised a sharp peak corresponding to $ν_{\rm NS}\sim$130 mHz X-ray pulsations along with a prominent quasi-periodic oscillation (QPO) feature at $ν_{\rm QPO}\sim$46 mHz with $\sim$20\% rms amplitude, which was positively correlated with energy. We also report the detection of sidebands to QPO occurring at a beat frequency ($ν_{\rm NS}-ν_{\rm QPO}$) of $\sim$83 mHz with $\sim$8\% rms amplitude, having $>3σ$ detection significance. Additionally, we utilized $Nuclear ~Spectroscopic ~Telescope ~ARray$ ($NuSTAR$) observations from the same torque state (2023 May-July) to analogize the presence and energy dependence of sidebands. The source retains timing properties in this spin-down torque state, similar to those seen in the previous spin-down phase. In sharp contrast, PDS from the 2018 observation was dominated by red noise, an absence of QPOs and a broadening in the wings of the pulse frequency peak, indicating a coupling between periodic and low-frequency aperiodic variability. Furthermore, we detected the known cyclotron resonance scattering feature (CRSF) at 37 keV in the Large Area X-ray Proportional Counter (LAXPC) spectrum. We explore various mechanisms that could possibly explain the presence of QPOs exclusively during the spin-down state.

astro-ph.HE↗

arXiv2024

Probing the energy and luminosity-dependent spectro-timing properties of RX J0440.9+4431 with AstroSat

The Be/X-ray binary pulsar RX J0440.9+4431 went through a giant outburst in December 2022 with a peak flux of $\sim$2.3 Crab in 15--50 keV. We studied the broad-band timing and spectral properties of RX J0440.9+4431 using four $AstroSat$ observations, where the source transited between subcritical and supercritical accretion regimes. Pulsations were detected significantly above 100 keV. The pulse profiles were found to be highly luminosity- and energy-dependent. A significant evolution in the pulse profile shape near the peak of the outburst indicates a possible change in the accretion mode and beaming patterns of RX J0440.9+4431. The rms pulsed fraction was luminosity- and energy-dependent, with a concave-like feature around 20--30 keV. The depth of this feature varied with luminosity, indicating changes in the accretion column height and proportion of reflected photons. The broad-band continuum spectra were best fitted with a two-component Comptonization model with a blackbody component or a two-blackbody component model with a thermal Comptonization component. A quasi-periodic oscillation at 60 mHz was detected at a luminosity of $2.6 \times 10^{37}$ erg s$^{-1}$, which evolved into 42 mHz at $1.5 \times 10^{37}$ erg s$^{-1}$. The QPO rms were found to be energy dependent with an overall increasing trend with energy. For the first time, we found the QPO frequency varying with photon energy in an X-ray pulsar, which poses a challenge in explaining the QPO with current models such as the Keplarian and beat frequency model. Hence, more physically motivated models are required to understand the physical mechanism behind the mHz QPOs.

astro-ph.HE↗

arXiv2024

A Prototype Model of Zero-Trust Architecture Blockchain with EigenTrust-Based Practical Byzantine Fault Tolerance Protocol to Manage Decentralized Clinical Trials

The COVID-19 pandemic necessitated the emergence of decentralized Clinical Trials (DCTs) due to patient retention, accelerate trials, improve data accessibility, enable virtual care, and facilitate seamless communication through integrated systems. However, integrating systems in DCTs exposes clinical data to potential security threats, making them susceptible to theft at any stage, a high risk of protocol deviations, and monitoring issues. To mitigate these challenges, blockchain technology serves as a secure framework, acting as a decentralized ledger, creating an immutable environment by establishing a zero-trust architecture, where data are deemed untrusted until verified. In combination with Internet of Things (IoT)-enabled wearable devices, blockchain secures the transfer of clinical trial data on private blockchains during DCT automation and operations. This paper proposes a prototype model of the Zero-Trust Architecture Blockchain (z-TAB) to integrate patient-generated clinical trial data during DCT operation management. The EigenTrust-based Practical Byzantine Fault Tolerance (T-PBFT) algorithm has been incorporated as a consensus protocol, leveraging Hyperledger Fabric. Furthermore, the Internet of Things (IoT) has been integrated to streamline data processing among stakeholders within the blockchain platforms. Rigorous evaluation has been done to evaluate the quality of the system.

cs.CR↗

arXiv2024

Physics-Informed Machine Learning for Smart Additive Manufacturing

Compared to physics-based computational manufacturing, data-driven models such as machine learning (ML) are alternative approaches to achieve smart manufacturing. However, the data-driven ML's "black box" nature has presented a challenge to interpreting its outcomes. On the other hand, governing physical laws are not effectively utilized to develop data-efficient ML algorithms. To leverage the advantages of ML and physical laws of advanced manufacturing, this paper focuses on the development of a physics-informed machine learning (PIML) model by integrating neural networks and physical laws to improve model accuracy, transparency, and generalization with case studies in laser metal deposition (LMD).

cs.LG↗

arXiv2024

TRUCE: Private Benchmarking to Prevent Contamination and Improve Comparative Evaluation of LLMs

Benchmarking is the de-facto standard for evaluating LLMs, due to its speed, replicability and low cost. However, recent work has pointed out that the majority of the open source benchmarks available today have been contaminated or leaked into LLMs, meaning that LLMs have access to test data during pretraining and/or fine-tuning. This raises serious concerns about the validity of benchmarking studies conducted so far and the future of evaluation using benchmarks. To solve this problem, we propose Private Benchmarking, a solution where test datasets are kept private and models are evaluated without revealing the test data to the model. We describe various scenarios (depending on the trust placed on model owners or dataset owners), and present solutions to avoid data contamination using private benchmarking. For scenarios where the model weights need to be kept private, we describe solutions from confidential computing and cryptography that can aid in private benchmarking. We build an end-to-end system, TRUCE, that enables such private benchmarking showing that the overheads introduced to protect models and benchmark are negligible (in the case of confidential computing) and tractable (when cryptographic security is required). Finally, we also discuss solutions to the problem of benchmark dataset auditing, to ensure that private benchmarks are of sufficiently high quality.

cs.CR↗

arXiv2024

Decoding Cognitive Health Using Machine Learning: A Comprehensive Evaluation for Diagnosis of Significant Memory Concern

The timely identification of significant memory concern (SMC) is crucial for proactive cognitive health management, especially in an aging population. Detecting SMC early enables timely intervention and personalized care, potentially slowing cognitive disorder progression. This study presents a state-of-the-art review followed by a comprehensive evaluation of machine learning models within the randomized neural networks (RNNs) and hyperplane-based classifiers (HbCs) family to investigate SMC diagnosis thoroughly. Utilizing the Alzheimer's Disease Neuroimaging Initiative 2 (ADNI2) dataset, 111 individuals with SMC and 111 healthy older adults are analyzed based on T1W magnetic resonance imaging (MRI) scans, extracting rich features. This analysis is based on baseline structural MRI (sMRI) scans, extracting rich features from gray matter (GM), white matter (WM), Jacobian determinant (JD), and cortical thickness (CT) measurements. In RNNs, deep random vector functional link (dRVFL) and ensemble dRVFL (edRVFL) emerge as the best classifiers in terms of performance metrics in the identification of SMC. In HbCs, Kernelized pinball general twin support vector machine (Pin-GTSVM-K) excels in CT and WM features, whereas Linear Pin-GTSVM (Pin-GTSVM-L) and Linear intuitionistic fuzzy TSVM (IFTSVM-L) performs well in the JD and GM features sets, respectively. This comprehensive evaluation emphasizes the critical role of feature selection and model choice in attaining an effective classifier for SMC diagnosis. The inclusion of statistical analyses further reinforces the credibility of the results, affirming the rigor of this analysis. The performance measures exhibit the suitability of this framework in aiding researchers with the automated and accurate assessment of SMC. The source codes of the algorithms and datasets used in this study are available at https://github.com/mtanveer1/SMC.

cs.LG↗

arXiv2024

Energy dependence of Quasi-periodic oscillations in accreting X-ray pulsars

We present the results from an investigation of the energy dependence of Quasi-Periodic Oscillations (QPOs) exhibited by accreting X-ray pulsars using data from archival \textit{XMM-Newton}, \textit{NuSTAR}, \textit{RXTE}, and \textit{NICER} observations. In a search for the presence of QPOs in 99 \textit{XMM-Newton} and \textit{NuSTAR} observations, we detected QPOs in eleven observations of five sources, viz., 4U 1626--67 (48 mHz), IGR J19294+1816 (30 mHz), V 0332+53 (2, 18 and 40 mHz), Cen X--3 (30 mHz), and XTE J1858+034 (180 mHz). A positive correlation of the QPO rms amplitude with energy is exhibited by 4U 1626--67, IGR J19294+1816, Cen X--3 and XTE J1858+034, while no energy dependence is observed in V 0332+53. We also analysed the energy spectrum to decouple thermal (soft-excess) from non-thermal emission and determine if the soft-excess has different QPO properties. We found no evidence for different QPO characteristics of the soft excess. The \textit{NuSTAR} observations of V 0332+53 during the Type-I outburst in 2016 show the presence of twin QPOs at 2.5 mHz and 18 mHz, while the \textit{XMM-Newton} and \textit{NuSTAR} observations during the Type-II outburst in 2015 show a QPO at 40 mHz. We review the observed QPO properties in the context of QPOs found in other types of accreting sources and the models usually used to explain the QPOs in accreting X-ray pulsars.

astro-ph.HE↗

arXiv2024

Enabling Memory Safety of C Programs using LLMs

Memory safety violations in low-level code, written in languages like C, continues to remain one of the major sources of software vulnerabilities. One method of removing such violations by construction is to port C code to a safe C dialect. Such dialects rely on programmer-supplied annotations to guarantee safety with minimal runtime overhead. This porting, however, is a manual process that imposes significant burden on the programmer and, hence, there has been limited adoption of this technique. The task of porting not only requires inferring annotations, but may also need refactoring/rewriting of the code to make it amenable to such annotations. In this paper, we use Large Language Models (LLMs) towards addressing both these concerns. We show how to harness LLM capabilities to do complex code reasoning as well as rewriting of large codebases. We also present a novel framework for whole-program transformations that leverages lightweight static analysis to break the transformation into smaller steps that can be carried out effectively by an LLM. We implement our ideas in a tool called MSA that targets the CheckedC dialect. We evaluate MSA on several micro-benchmarks, as well as real-world code ranging up to 20K lines of code. We showcase superior performance compared to a vanilla LLM baseline, as well as demonstrate improvement over a state-of-the-art symbolic (non-LLM) technique.

cs.SE↗

arXiv2024

A comprehensive study of orbital evolution of LMC X-4: Existence of a second derivative of the orbital period

We report here results from pulse arrival time delay analysis of the eclipsing high mass X-ray binary pulsar LMC X-4 using observations made with the Rossi X-ray Timing Explorer, XMM-Newton, NuSTAR and AstroSat. Combining the orbital parameters determined from these observations with the historical measurements dating back to 1998, we have extended the $T_{π/2}$ epoch history of LMC X-4 by about 4600 binary orbits spanning about 18 years. We also report mid-eclipse time measurements ($T_{ecl}$) using data obtained from wide-field X-ray monitors of MAXI-GSC and Swift-BAT. Combining the new $T_{π/2}$ and $T_{ecl}$ estimates with all the previously reported values, we have significantly improved the orbital evolution measurement, which indicates that the orbital period is evolving at a time scale ($P_{\rm orb}/\dot{P}_{\rm orb}$ ) of about 0.8 Myr. For the first time in an accreting X-ray pulsar system, we confirm the existence of a second derivative of the orbital period, having an evolution time scale ($\dot{P}_{orb}/\ddot{P}_{orb}$) of about 55 yr. Detection of a second derivative of the orbital period in LMC X-4 makes its orbital evolution timescale more uncertain, which may also be true for other HMXBs. Independent solutions for the orbital evolution measurement using the mid-eclipse data and the pulse timing data are consistent with each other, and help us put an upper limit of 0.009 on the eccentricity of the binary system.

astro-ph.HE↗

arXiv2024

Flares during Eclipses of High Mass X-ray Binary Systems Vela X-1, 4U 1700-37, and LMC X-4

In eclipsing X-ray binary systems, the direct X-ray emission is blocked by the companion star during the eclipse. We observe only reprocessed emission that contains clues about the environment of the compact object and its chemical composition, ionization levels, etc. We have found flares in some X-ray binaries during their eclipses. The study of eclipse flares provides additional clues regarding the size of the reprocessing region and helps distinguish between different components of the X-ray spectrum observed during the eclipse. In the archival data, we searched for flares during eclipses of high-mass X-ray binaries and found flares in three sources: Vela X-1, LMC X-4, and 4U 1700-37. Comparing spectral properties of the eclipse flare and non-flare data, we found changes in the power-law photon index in all three sources and multiple emission lines in Vela X-1 and 4U 1700-37. The fluxes of prominent emission lines showed a similar increase as the overall X-ray flux during the eclipse flare, suggesting the lines originate in the binary environment and not in the interstellar medium. We also observed a soft excess in 4U 1700-37 that remains unchanged during both eclipse flare and non-flare states. Our analysis suggests that this emission originates from the extremely thin shell of the stellar wind surrounding the photosphere of its companion star. The detection of short (100-200 seconds) count-rate doubling timescale in 4U 1700-37 and LMC X-4 indicates that the eclipse reprocessing occurs in a region larger than, but comparable to the size of the companion star.

astro-ph.HE↗

arXiv2024

Detection of a cyclotron line in the Be X-ray pulsar IGR J06074$+$2205

IGR J0607.4+2205 is a transient Be X-ray binary discovered two decades ago. IGR J0607.4+2205 underwent an outburst in 2023 during which it was observed twice with \textit{NuSTAR}. The main goal of this work is to model the broadband X-ray spectrum of IGR J0607.4+2205 during the outburst and to study the variations of the spectral and timing features at different intensities. We extracted the light curve and spectrum of the source from the two \textit{NuSTAR} observations carried out during the recent outburst in the energy range of 3$-$78 keV. We used the epoch folding technique to find pulsation from the source and to study the changes in emission characteristics from the source with energy across an order of magnitude variation in source luminosity. IGR J0607.4+2205 shows pulsations with a period of $\sim$347.6 s during both the observations, with a pulse fraction of $\geq$50\%. The broadband spectrum of the source was modelled using a power-law continuum with a high-energy cutoff. During the first observation, a cyclotron absorption line at $\sim$51 keV was also present in the source with an optical depth of $\sim$1.3. However, no cyclotron line feature was detected in the second observation when the source was an order of magnitude fainter. Additionally, soft excess was detected in the second observation, which was modelled with a black body component emerging from close to the neutron star (NS). We report the first ever detection of a cyclotron line in the broadband spectrum of IGR J0607.4+2205 centred at 51$\pm$1 keV. The magnetic field strength of the NS is estimated to be $\sim$4$\times$$10^{12}$ G from the centroid energy of the absorption line. A significant change is observed in the pulse profile with luminosity during the decay of the outburst, indicating an associated change in the beaming pattern.

astro-ph.HE↗

arXiv2024

Evaluating Atypical Gaze Patterns through Vision Models: The Case of Cortical Visual Impairment

A wide range of neurological and cognitive disorders exhibit distinct behavioral markers aside from their clinical manifestations. Cortical Visual Impairment (CVI) is a prime example of such conditions, resulting from damage to visual pathways in the brain, and adversely impacting low- and high-level visual function. The characteristics impacted by CVI are primarily described qualitatively, challenging the establishment of an objective, evidence-based measure of CVI severity. To study those characteristics, we propose to create visual saliency maps by adequately prompting deep vision models with attributes of clinical interest. After extracting saliency maps for a curated set of stimuli, we evaluate fixation traces on those from children with CVI through eye tracking technology. Our experiments reveal significant gaze markers that verify clinical knowledge and yield nuanced discriminability when compared to those of age-matched control subjects. Using deep learning to unveil atypical visual saliency is an important step toward establishing an eye-tracking signature for severe neurodevelopmental disorders, like CVI.

eess.SP↗

arXiv2024

Ranking LLM-Generated Loop Invariants for Program Verification

Synthesizing inductive loop invariants is fundamental to automating program verification. In this work, we observe that Large Language Models (such as gpt-3.5 or gpt-4) are capable of synthesizing loop invariants for a class of programs in a 0-shot setting, yet require several samples to generate the correct invariants. This can lead to a large number of calls to a program verifier to establish an invariant. To address this issue, we propose a {\it re-ranking} approach for the generated results of LLMs. We have designed a ranker that can distinguish between correct inductive invariants and incorrect attempts based on the problem definition. The ranker is optimized as a contrastive ranker. Experimental results demonstrate that this re-ranking mechanism significantly improves the ranking of correct invariants among the generated candidates, leading to a notable reduction in the number of calls to a verifier. The source code and the experimental data for this paper are available in \url{https://github.com/microsoft/NeuralInvariantRanker}.

cs.PL↗

arXiv2023

$\textit{Insights}$ into the phase-dependent cyclotron line feature in XTE J1946+274: An $\textit{AstroSat}$ and $\textit{Insight}$-HXMT view

XTE J1946+274 is a Be/X-ray binary with a 15.8s spin period and 172 d orbital period. Using $\textit{RXTE/PCA}$ data of the 1998 outburst, a cyclotron line around 37 keV was reported. The presence of this line, its dependence on the pulse phase, and its variation with luminosity have been of some debate since. In this work, we present the reanalysis of two $\textit{AstroSat}$ observations: one made during the rising phase of the 2018 outburst and the other during the declining phase of the 2021 outburst. We also present a new analysis of the $\textit{Insight}$-HXMT observations of the source at the peak of the 2018 outburst. We find the source to be spinning up over the course of the outburst and spinning down between the two outbursts. We report the presence of a higher cyclotron line energy using the 2018 $\textit{AstroSat}$ observation ($\sim 45$ keV) and 2018 $\textit{Insight}$-HXMT observation ($\sim$ 50 keV) and a line at $\sim$ 40 keV during the declining phase of the 2021 outburst using data from $\textit{AstroSat}$. We also investigate the pulse phase dependence of the cyclotron line parameters and find that the line is significantly detected in all the phases of both $\textit{AstroSat}$ observations, along with showing variation with the pulse phase. This differs from the previous results reported using $\textit{BeppoSAX}$ and $\textit{NuSTAR}$. We explain this behaviour of the cyclotron line to be due to photon spawning and different accretion column radii at the two poles of this neutron star.

astro-ph.HE↗

arXiv2023

Development and Validation of a Dynamic Operating Envelopes-enabled Demand Response Scheme in Low-voltage Distribution Networks

Dynamic operating envelopes (DOEs) offer an attractive solution for maintaining network integrity amidst increasing penetration of distributed energy resources (DERs) in low-voltage (LV) networks. Currently, the focus of DOEs primarily revolves around active power exports of rooftop photovoltaic (PV) generation, often neglecting the impact of demand response (DR). This paper presents a two-stage, coordinated approach for residential DR participation in electricity markets under the DOE framework. In the first stage, the distribution network service provider (DNSP) adopts a convex hull technique to establish DOEs at each customer point-of-connection (POC). In the second stage, the demand response aggregator (DRA) utilises DOEs assigned by the DNSP to develop a hierarchical control scheme for tracking a load set-point signal without jeopardising network statutory limits. To assess the effectiveness of the proposed control scheme in a practical setting, software-in-the-loop (SIL) tests are performed in a grid simulator, considering a real residential feeder with realistic household load and generation profiles. Simulation validations suggest that the DRA can provide precise DR while honouring network statutory limits and maintaining end-user thermal comfort. Furthermore, the overall approach is compliant with the market dispatch interval and preserves end-user data privacy.

eess.SY↗

arXiv2023

Finding Inductive Loop Invariants using Large Language Models

Loop invariants are fundamental to reasoning about programs with loops. They establish properties about a given loop's behavior. When they additionally are inductive, they become useful for the task of formal verification that seeks to establish strong mathematical guarantees about program's runtime behavior. The inductiveness ensures that the invariants can be checked locally without consulting the entire program, thus are indispensable artifacts in a formal proof of correctness. Finding inductive loop invariants is an undecidable problem, and despite a long history of research towards practical solutions, it remains far from a solved problem. This paper investigates the capabilities of the Large Language Models (LLMs) in offering a new solution towards this old, yet important problem. To that end, we first curate a dataset of verification problems on programs with loops. Next, we design a prompt for exploiting LLMs, obtaining inductive loop invariants, that are checked for correctness using sound symbolic tools. Finally, we explore the effectiveness of using an efficient combination of a symbolic tool and an LLM on our dataset and compare it against a purely symbolic baseline. Our results demonstrate that LLMs can help improve the state-of-the-art in automated program verification.

cs.PL↗