arXiv subjects

Prashant Singh

Publications and source records attributed to Prashant Singh.

At least 37 records · Page 2Linked to original sources

arXiv2025

Crystal to liquid cross-over for active particles with inverse-square power-law interaction

We consider a one-dimensional system comprising of $N$ run-and-tumble particles confined in a harmonic trap interacting via a repulsive inverse-square power-law interaction. We numerically compute the global density profile in the steady state which shows interesting crossovers between three different regimes: as the activity increases, we observe a change from a density with sharp peaks characteristic of a crystal region to a smooth bell-shaped density profile, passing through the intermediate stage of a smooth Wigner semi-circle characteristic of a liquid phase. We also investigate analytically the crossover between the crystal and the liquid regions by computing the covariance of the positions of these particles in the steady state in the weak noise limit. It is achieved by using the method introduced in Touzo {\it et al.} [Phys. Rev. E {\bf 109}, 014136 (2024)] to study the active Dyson Brownian motion. Our analytical results are corroborated by thorough numerical simulations.

cond-mat.stat-mech↗

arXiv2025

Bayesian Robust Aggregation for Federated Learning

Federated Learning enables collaborative training of machine learning models on decentralized data. This scheme, however, is vulnerable to adversarial attacks, when some of the clients submit corrupted model updates. In real-world scenarios, the total number of compromised clients is typically unknown, with the extent of attacks potentially varying over time. To address these challenges, we propose an adaptive approach for robust aggregation of model updates based on Bayesian inference. The mean update is defined by the maximum of the likelihood marginalized over probabilities of each client to be `honest'. As a result, the method shares the simplicity of the classical average estimators (e.g., sample mean or geometric median), being independent of the number of compromised clients. At the same time, it is as effective against attacks as methods specifically tailored to Federated Learning, such as Krum. We compare our approach with other aggregation schemes in federated setting on three benchmark image classification data sets. The proposed method consistently achieves state-of-the-art performance across various attack types with static and varying number of malicious clients.

cs.LG↗

arXiv2025

Fermi surface nesting driven anomalous Hall effect in magnetically frustrated Mn_2PdIn

Noncollinear magnets with near-zero net magnetization and nontrivial bulk electronic topology hold significant promise for spintronic applications, though their scarcity necessitates purposeful design strategies. In this work, we report a topologically nontrivial electronic structure in metallic Mn_2PdIn, which crystallizes in the inverse Heusler structure and exhibits a spin-glassy ground state with quenched magnetization. The system features Weyl-type band crossings near the Fermi level and reveals a novel interplay among momentum-space nesting, orbital hybridization, and spin-orbit coupling. Comprehensive transport measurements uncover a pronounced anomalous Hall effect (AHE) in Mn_2PdIn. The observed quadratic relationship between the longitudinal and anomalous Hall resistivities highlights the intrinsic Berry curvature contribution to AHE. These findings establish inverse Heusler alloys as compelling platforms for realizing noncollinear magnets that host Weyl-type semimetallic or metallic phases-combining suppressed magnetization with robust electronic transport-thereby offering a promising route toward their seamless integration into next-generation spintronic devices.

cond-mat.mtrl-sci↗

arXiv2025

Interpretable machine learning-guided design of Fe-based soft magnetic alloys

We present a machine-learning guided approach to predict saturation magnetization (MS) and coercivity (HC) in Fe-rich soft magnetic alloys, particularly Fe-Si-B systems. ML models trained on experimental data reveals that increasing Si and B content reduces MS from 1.81T (DFT~2.04 T) to ~1.54 T (DFT~1.56T) in Fe-Si-B, which is attributed to decreased magnetic density and structural modifications. Experimental validation of ML predicted magnetic saturation on Fe-1Si-1B (2.09T), Fe-5Si-5B (2.01T) and Fe-10Si-10B (1.54T) alloy compositions further support our findings. These trends are consistent with density functional theory (DFT) predictions, which link increased electronic disorder and band broadening to lower MS values. Experimental validation on selected alloys confirms the predictive accuracy of the ML model, with good agreement across compositions. Beyond predictive accuracy, detailed uncertainty quantification and model interpretability including through feature importance and partial dependence analysis reveals that MS is governed by a nonlinear interplay between Fe content, early transition metal ratios, and annealing temperature, while HC is more sensitive to processing conditions such as ribbon thickness and thermal treatment windows. The ML framework was further applied to Fe-Si-B/Cr/Cu/Zr/Nb alloys in a pseudo-quaternary compositional space, which shows comparable magnetic properties to NANOMET (Fe84.8Si0.5B9.4Cu0.8 P3.5C1), FINEMET (Fe73.5Si13.5B9 Cu1Nb3), NANOPERM (Fe88Zr7B4Cu1), and HITPERM (Fe44Co44Zr7B4Cu1. Our fundings demonstrate the potential of ML framework for accelerated search of high-performance, Co- and Ni-free, soft magnetic materials.

cond-mat.mtrl-sci↗

arXiv2025

Effects of chemical disorder and spin-orbit coupling on electronic-structure and Fermi-surface topology of YbSb-based monopnictides

In this work, we study the influence of disorder on the electronic structure of YbSb -- a rare-earth monopnictide featuring a simple rock-salt (B1) crystal structure and a well-defined Fermi surface topology -- by employing first-principles density functional theory (DFT). We focus on chemical disorder introduced through Te and Al doping, selected based on their thermodynamic stability in alloyed configurations, to understand how such perturbations modify the electronic states of YbSb. Our results indicate that Te doping predominantly introduces electron-like states at the \textit{X} and \textit{L} points, while Al doping leads to a suppression of hole-like states at $Γ$, effectively driving the system from a semimetallic state to one characterized by very narrow-gap behavior at $Γ$. This modulation of the Fermi surface, particularly the reduction of central hole pockets at $Γ$, plays a central role in altering inter-pocket scattering -- a mechanism critical for tuning quantum transport properties, including superconductivity. This disorder-driven modulation of the Fermi surface, particularly the suppression of central hole pockets at $Γ$, controls inter-pocket scattering, which is essential for optimizing quantum transport properties, including superconductivity. Our results show that disorder can be effectively used as a means of engineering band topology, thereby tuning quantum-related responses through tailored electronic structure.

cond-mat.mtrl-sci↗

arXiv2025

PARIC: Probabilistic Attention Regularization for Language Guided Image Classification from Pre-trained Vison Language Models

Language-guided attention frameworks have significantly enhanced both interpretability and performance in image classification; however, the reliance on deterministic embeddings from pre-trained vision-language foundation models to generate reference attention maps frequently overlooks the intrinsic multivaluedness and ill-posed characteristics of cross-modal mappings. To address these limitations, we introduce PARIC, a probabilistic framework for guiding visual attention via language specifications. Our approach enables pre-trained vision-language models to generate probabilistic reference attention maps, which align textual and visual modalities more effectively while incorporating uncertainty estimates, as compared to their deterministic counterparts. Experiments on benchmark test problems demonstrate that PARIC enhances prediction accuracy, mitigates bias, ensures consistent predictions, and improves robustness across various datasets.

cs.CV↗

arXiv2025

Machine learning driven search of hydrogen storage materials

The transition to a low-carbon economy demands efficient and sustainable energy-storage solutions, with hydrogen emerging as a promising clean-energy carrier and with metal hydrides recognized for their hydrogen-storage capacity. Here, we leverage machine learning (ML) to predict hydrogen-to-metal (H/M) ratios and solution energy by incorporating thermodynamic parameters and local lattice distortion (LLD) as key features. Our best-performing ML model provides improvements to H/M ratios and solution energies over a broad class of ternary alloys (easily extendable to multi-principal-element alloys), such as Ti-Nb-X (X = Mo, Cr, Hf, Ta, V, Zr) and Co-Ni-X (X = Al, Mg, V). Ti-Nb-Mo alloys reveal compositional effects in H-storage behavior, in particular Ti, Nb, and V enhance H-storage capacity, while Mo reduces H/M and hydrogen weight percent by 40-50%. We attributed to slow hydrogen kinetics in molybdenum rich alloys, which is validated by our pressure-composition isotherm (PCT) experiments on pure Ti and Ti5Mo95 alloys. Density functional theory (DFT) and molecular simulations also confirm that Ti and Nb promote H diffusion, whereas Mo hinders it, highlighting the interplay between electronic structure, lattice distortions, and hydrogen uptake. Notably, our Gradient Boosting Regression model identifies LLD as a critical factor in H/M predictions. To aid material selection, we present two periodic tables illustrating elemental effects on (a) H2 wt% and (b) solution energy, derived from ML, and provide a reference for identifying alloying elements that enhance hydrogen solubility and storage.

cond-mat.mtrl-sci↗

arXiv2025

Linking quantum mechanical features to structural phase-transformation in inorganic solids

We present a new descriptor, i.e., local lattice distortion, to predict structural phase transformation in inorganic compounds containing lanthanides and transition metals. The descriptor utilizes local lattice and angular distortions obtained from structural optimization of experimentally known crystalline phases within state-of-the-art density-functional theory method. The predictive power of the descriptor was tested on lanthanide based RE2In (RE=rare-earth) compounds known for a variety of phase transformations. We show that the inclusion of quantum-mechanical effects through local-charge, bonding, symmetry, and electronic-structure enhances the robustness of the descriptor in predicting structural phase transformation. To gain further insights, we analyzed phononic and electronic behavior of Y2In, and show that experimentally observed phase transformation can only be predicted when atomic strains are included. The descriptor was used to predict structural phase change in couple of new compounds, i.e., (Yb1-xErx)2In and Gd2(In1-xAlx), which was validated by X-ray powder diffraction measurements. Finally, we demonstrated the generality of the proposed descriptor by predicting phase transformation behavior in different classes of compounds indicating the usefulness of our approach in mapping desired phase changes in novel functional materials.

cond-mat.mtrl-sci↗

arXiv2025

Limits to positional information in boundary-driven systems

Chemical gradients can be used by a particle to determine its position. This \textit{positional information} is of crucial importance, for example in developmental biology in the formation of patterns in an embryo. The central goal of this paper is to study the fundamental physical limits on how much positional information can be stored inside a system. To achieve this, we study positional information for general boundary-driven systems, and derive, in the near-equilibrium regime, a universal expression involving only the chemical potential and density gradients of the system. We also conjecture that this expression serves as an upper bound on the positional information of boundary driven systems beyond linear response. To support this claim, we test it on a broad range of solvable boundary-driven systems.

cond-mat.stat-mech↗

arXiv2025

Charge Order Driven Multiferroic Behaviour in Sr$_4$Fe$_6$O$_{12}$: An $\textit{Ab-initio}$ Study

In this letter, we report the structural, electronic and ferroelectric properties of the layered mixed-valent transition-metal compound, Sr$_{4}$Fe$_{6}$O$_{12}$ (SFO). We demonstrate how SFO undergoes a phase transition from a high-temperature (T) centrosymmetric tetragonal phase ($P4_{2}/mnm$) to a low-T polar orthorhombic phase ($Pmn2_{1}$). The transition is primarily driven by charge ordering at tetrahedral Fe-layer creating Fe$^{3+}$ and Fe$^{2+}$ cations between two edge sharing tetrahedra. This charge ordering induces electronic polarization, which is remarkably larger (3.5 times) in magnitude than ionic polarization and oppositely directed, giving a net polarization of 0.05 C/m$^2$ which is comparable to the state-of-the-art rare-earth nickelets and manganite perovskites. The direction of structural distortion, governed by the polar mode irrep $Γ_{5}^{-}$, depends sensitively on the type of magnetic ordering in the Fe-octahedral layer. Consequently, both the ionic and electronic polarization directions are influenced by magnetic ordering, suggesting the potential for multiferroic behavior with strong magneto-electric coupling in this material.

cond-mat.str-el↗

arXiv2025

$\texttt{InfoHier}$: Hierarchical Information Extraction via Encoding and Embedding

Analyzing large-scale datasets, especially involving complex and high-dimensional data like images, is particularly challenging. While self-supervised learning (SSL) has proven effective for learning representations from unlabelled data, it typically focuses on flat, non-hierarchical structures, missing the multi-level relationships present in many real-world datasets. Hierarchical clustering (HC) can uncover these relationships by organizing data into a tree-like structure, but it often relies on rigid similarity metrics that struggle to capture the complexity of diverse data types. To address these we envision $\texttt{InfoHier}$, a framework that combines SSL with HC to jointly learn robust latent representations and hierarchical structures. This approach leverages SSL to provide adaptive representations, enhancing HC's ability to capture complex patterns. Simultaneously, it integrates HC loss to refine SSL training, resulting in representations that are more attuned to the underlying information hierarchy. $\texttt{InfoHier}$ has the potential to improve the expressiveness and performance of both clustering and representation learning, offering significant benefits for data analysis, management, and information retrieval.

cs.IR↗

arXiv2024

Positional information trade-offs in boundary-driven reaction-diffusion systems

Individual components such as cells, particles, or agents within a larger system often require detailed understanding of their relative position to act accordingly, enabling the system as a whole to function in an organised and efficient manner. Through the concept of positional information, such components are able to specify their position in order to, e.g., create robust spatial patterns or coordinate specific functionality. Such complex behaviour generally occurs far from thermodynamic equilibrium and thus requires the dissipation of free energy to sustain functionality. We show that in boundary-driven simple exclusion systems with position-dependent Langmuir kinetics, non-trivial Pareto-optimal trade-offs exist between the positional information, rescaled entropy production rate and global reaction current. Phase transitions in the optimal protocols that tune the densities of the system boundaries emerge as a result, showing that distinct protocols are able to exchange global optimality similar to phase coexistence in liquid-gas phase transitions, and that increasing the positional information can lead to diminishing returns when considering increased dissipation.

cond-mat.stat-mech↗

arXiv2024

The influence of Ga doping on magnetic properties, magnetocaloric effect, and electronic structure of pseudo-binary GdZn1-xGax (x = 0-0.1)

We explore the impact of introducing IIIA-group element Ga in place of IIB-group element Zn in binary intermetallic GdZn on its magnetic and magnetocaloric properties, as well as explicate the modified electronic band structure of the compound. The magnetic transition temperature of the compound decreases with the increase of Ga concentration in GdZn1-xGax (x = 0-0.1) while the crystal structure (CsCl-prototype) and lattice parameters remain unchanged. Our detailed analysis of magnetization and magnetocaloric data conclusively proves that long-ranged magnetic ordering exists in the sample, despite the magnetic interaction considerably weakening with the increase of Ga. The experimental data is rationalized using both theoretical machine learning model and first-principle density functional theory.The electronic band structure of GdZn is manifested with some unusual complex features which gradually diminish with Ga doping and conventional sinusoidal feature of Ruderman-Kittel-Kasuya- Yosida (RKKY)-type interactions also disappears. A mean-field theory model is developed and can successfully describe the overall magnetocaloric behavior of the GdZn1-xGax series of samples

cond-mat.mtrl-sci↗

arXiv2024

Emerging cost-time Pareto front for diffusion with stochastic return

Resetting, in which a system is regularly returned to a given state after a fixed or random duration, has become a useful strategy to optimize the search performance of a system. While earlier theoretical frameworks focused on instantaneous resetting, wherein the system is directly teleported to a given state, there is a growing interest in physical resetting mechanisms that involve a finite return time. However employing such a mechanism involves cost and the effect of this cost on the search time remains largely unexplored. Yet answering this is important in order to design cost-efficient resetting strategies. Motivated from this, we present a thermodynamic analysis of a diffusing particle whose position is intermittently reset to a specific site by employing a stochastic return protocol with external confining trap. We show for a family of potentials $U_R(x) \sim |x|^{m}$ with $m>0$, it is possible to find optimal potential shape that minimises the expected first-passage time for a given value of the thermodynamic cost, i.e. mean work. By varying this value, we then obtain the Pareto optimal front, and demonstrate a trade-off relation between the first-passage time and the work done.

cond-mat.stat-mech↗

arXiv2024

EnterpriseEM: Fine-tuned Embeddings for Enterprise Semantic Search

Enterprises grapple with the significant challenge of managing proprietary unstructured data, hindering efficient information retrieval. This has led to the emergence of AI-driven information retrieval solutions, designed to adeptly extract relevant insights to address employee inquiries. These solutions often leverage pre-trained embedding models and generative models as foundational components. While pre-trained embeddings may exhibit proximity or disparity based on their original training objectives, they might not fully align with the unique characteristics of enterprise-specific data, leading to suboptimal alignment with the retrieval goals of enterprise environments. In this paper, we propose a comprehensive methodology for contextualizing pre-trained embedding models to enterprise environments, covering the entire process from data preparation to model fine-tuning and evaluation. By adapting the embeddings to better suit the retrieval tasks prevalent in enterprises, we aim to enhance the performance of information retrieval solutions. We discuss the process of fine-tuning, its effect on retrieval accuracy, and the potential benefits for enterprise information management. Our findings demonstrate the efficacy of fine-tuned embedding models in improving the precision and relevance of search results in enterprise settings.

cs.IR↗

arXiv2024

Inferring entropy production from time-dependent moments

Measuring entropy production of a system directly from the experimental data is highly desirable since it gives a quantifiable measure of the time-irreversibility for non-equilibrium systems and can be used as a cost function to optimize the performance of the system. Although numerous methods are available to infer the entropy production of stationary systems, there are only a limited number of methods that have been proposed for time-dependent systems and, to the best of our knowledge, none of these methods have been applied to experimental systems. Herein, we develop a general non-invasive methodology to infer a lower bound on the mean total entropy production for arbitrary time-dependent continuous-state Markov systems in terms of the moments of the underlying state variables. The method gives surprisingly accurate estimates for the entropy production, both for theoretical toy models and for experimental bit erasure, even with a very limited amount of experimental data.

cond-mat.stat-mech↗

arXiv2024

Tracer dynamics in the active random average process

We investigate the dynamics of tracer particles in the random average process (RAP), a single-file system in one dimension. In addition to the position, every particle possesses an internal spin variable $σ(t)$ that can alternate between two values, $\pm 1$, at a constant rate $γ$. Physically, the value of $σ(t)$ dictates the direction of motion of the corresponding particle and for finite $γ$, every particle performs a non-Markovian active dynamics. Herein, we study the effect of this non-Markovianity in the fluctuations and correlations of the positions of tracer particles. We analytically show that the variance of the position of a tagged particle grows sub-diffusively as $\sim ζ_{\text{q}} \sqrt{t}$ at large times for the quenched uniform initial condition. While this sub-diffusive growth is identical to that of the Markovian/non-persistent RAP, the coefficient $ζ_{\text{q}} $ is rather different and bears the signature of the persistent motion of active particles through higher point correlations (unlike in the Markovian case). Similarly, for the annealed (steady state) initial condition, we find that the variance scales as $\sim ζ_{\text{a}} \sqrt{t}$ at large times with coefficient $ζ_{\text{a}} $ once again different from the non-persistent case. Although $ζ_{\text{q}}$ and $ζ_{\text{a}} $ both individually depart from their Markov counterparts, their ratio $ζ_{\text{a}} / ζ_{\text{q}}$ is still equal to $\sqrt{2}$, a condition observed for other diffusive single-file systems. This condition turns out to be true even in the strongly active regimes as corroborated by extensive simulations and calculations. Finally, we study the correlation between the positions of two tagged particles in both quenched uniform and annealed initial conditions. We verify all our analytic results by extensive numerical simulations.

cond-mat.stat-mech↗

arXiv2024

Adaptive Robust Learning using Latent Bernoulli Variables

We present an adaptive approach for robust learning from corrupted training sets. We identify corrupted and non-corrupted samples with latent Bernoulli variables and thus formulate the learning problem as maximization of the likelihood where latent variables are marginalized. The resulting problem is solved via variational inference, using an efficient Expectation-Maximization based method. The proposed approach improves over the state-of-the-art by automatically inferring the corruption level, while adding minimal computational overhead. We demonstrate our robust learning method and its parameter-free nature on a wide variety of machine learning tasks including online learning and deep learning where it adapts to different levels of noise and maintains high prediction accuracy.

stat.ML↗