arXiv subjects

Vipin Kumar

Publications and source records attributed to Vipin Kumar.

At least 19 recordsLinked to original sources

arXiv2026

Catching the Imposter: Self-Supervised Learning of Physical Coherence with Cross-Entity Feature Permutations

Scientific data often describe entities whose features are jointly governed by the laws of physics, yet existing self-supervised learning (SSL) objectives largely ignore this physical coherence. We introduce imposter, a discriminative pretext task that replaces subsets of an entity's features with real observations donated by another entity and trains the encoder to identify the swapped features. Because every donated value is individually plausible, the task can only be solved by learning cross-feature physical dependencies. We evaluate the proposed objectives on global ERA5-Land reanalysis data using 21 environmental variables and assess the learned representations on seven downstream tasks spanning climate classification, carbon flux estimation, and streamflow prediction. Our study includes, to our knowledge, the first systematic comparison of self-supervised objectives for land-surface modeling under a shared architecture and pre-training budget. We find that the most effective pretext task depends on the downstream task family rather than any single objective's superiority, and that imposter provides complementary information when combined with existing SSL objectives. These results suggest that physical coherence is a valuable new source of self-supervision for scientific foundation models.

cs.LG↗

arXiv2026

UAV3DCrop: Benchmarking 3D Reconstruction in Repeated Multi-Angle UAV Crop Surveys

Accurate 3D crop monitoring underpins data-driven precision agriculture by enabling field-scale analysis of plant structure, growth dynamics, and management response. Modern 3D reconstruction methods perform strongly on generic benchmarks, but rendered appearance may not translate into metrically and agronomically useful geometry in crop fields. We introduce UAV3DCrop, a public benchmark of repeated multi-angle unmanned aerial vehicle (UAV) crop surveys. It contains 88,830 RGB images at $5280 \times 3956$ pixels, with a ground sampling distance of 3.6-5.8 mm, from 91 scenes spanning corn, soybean, wheat, and oat. Track A evaluates seven scene-optimized methods -- Neural Radiance Field (NeRF) and 3D Gaussian Splatting (3DGS) variants -- on held-out views, photogrammetry-referenced depth, and canopy-height recovery. Track B tests four pretrained feed-forward models on zero-shot camera-pose and geometry estimation. The scene-optimized methods rank differently across the three targets: Splatfacto-big leads appearance, whereas Scaffold-GS leads depth and is statistically tied with Splatfacto for canopy height. Among feed-forward models, MapAnything leads on seven of the eight metrics, while the remaining models vary more across crops and fail severely on absolute scale in a way that alignment conceals. Repeated acquisitions reveal further sensitivities that differ by output type and by model, associated with position within the acquisition sequence and with tie-point multiplicity. Current 3D reconstruction methods are therefore not yet interchangeable for agronomic use: no single method wins on appearance, geometry, and canopy height at once, and only one of four feed-forward models recovers usable metric scale. The dataset is publicly available at https://link-dev.github.io/UAV3DCrop/

cs.CV↗

arXiv2026

Sutra : An integrated framework for identification and characterization of filaments in the interstellar medium

Observations of the interstellar medium (ISM) at Far-infrared(FIR) and sub-millimetre (sub-mm) wavelengths reveal a complex filamentary structure of dust and gas, which plays a pivotal role in both low and high mass star formation. Large scale identification and characterization of filaments is crucial to establish a link between the ISM and the star formation. We present Sutra, a machine learning based framework that unifies filament identification and beam-scale physical characterization within a single automated pipeline. The framework employs a U-Net architecture to perform supervised segmentation on column density maps and is trained on five nearby (<500pc) molecular clouds from the Herschel Gould Belt Survey (HGBS), using consensus skeletons constructed from the union of filaments identified by DisPerSE and getsf. Rather than reproducing broad intensity-based masks, Sutra predicts crest-likelihood maps focused on filament spines. Beyond identification, Sutra characterizes the filaments at the beam resolution by extracting radial profiles perpendicular to the crest and deriving local structural parameters. The framework provides a parameter-free, computationally efficient approach for consistent filaments identification and systematic investigation of their local properties and shows stable behaviour across varying background conditions in controlled synthetic tests. We demonstrate its application on selected regions from Aquila, Orion and Polaris molecular clouds, and compare the derived filament characteristics with those obtained using existing algorithms. Sutra robustly recovers filamentary structures consistent with cylindrical profiles, including in relatively low-intensity and low-contrast environments, making it well suited for both region-specific studies and large-scale statistical analyses of early-stage star formation and ISM structure.

astro-ph.GA↗

arXiv2026

Physics Informed Neural Networks for Nonlinear Delay Differential Equations

In this paper we propose a novel physics-informed neural network framework for solving general first-order delay differential equations. Our approach combines a differentiable history switch, a trial-solution formulation that explicitly enforces history constraints, and a segmented collocation strategy to stabilize gradient propagation across large temporal domains. The method enables a scalable and physics-consistent approximation of delay differential equation solutions while maintaining continuity across subintervals. Numerical experiments demonstrate the effectiveness of the proposed method.

math.NA↗

arXiv2026

On Surrogate Modeling of Static Response of AM Short-Fiber Thermoplastics Using Graph Neural Networks

Short-fiber thermoplastic (SFT) composites are increasingly employed in lightweight aerospace and automotive structures owing to their favorable strength-to-weight ratio, high production rates, and recyclability. Unlike continuous-fiber systems, the mechanical response of SFTs is governed by mesoscale interactions among fiber orientation, spatial clustering, and manufacturing-induced porosity. These features exhibit significant spatial variability in manufactured components and influence stiffness, damage initiation, and nonlinear deformation. Although mesoscale finite element (FE) models can resolve such heterogeneity, their application to realistic three-dimensional microstructures remains computationally intractable. A data-driven surrogate framework is proposed to predict the mechanical behavior of additively manufactured, compression-molded (AM-CM) SFTs. Microstructures reconstructed from micro-computed tomography data were discretized into Voronoi-based cells representing distinct fiber-interaction neighborhoods. Each cell was homogenized via nonlinear FE simulations incorporating matrix damage, and the resulting stress-strain responses trained a hybrid Graph Neural Network-Long Short-Term Memory (GNN-LSTM) architecture encoding microstructural topology and history-dependent mechanical evolution. The surrogate accurately predicts stiffness and stress-strain behavior of unseen microstructures, achieving $R^2\approx 0.98$ relative to high-fidelity FE simulations with over two orders-of-magnitude reduction in computational cost. Coupling the framework with experimentally calibrated damage laws demonstrates that fiber orientation, clustering, and porosity collectively govern local effective stiffness. The approach provides a physics-informed, data-efficient pathway to identify mechanically weak microstructural cells and accelerate digital-twin development for SFT components.

cs.LG↗

arXiv2026

Discovery of variable polarization in H$\alpha$ profile of symbiotic star Y Gem: A case for orbital-phase dependent variation of Raman-scattered Ly$\beta$ emission

The geometry and morphology of symbiotic stars are conducive to exhibit a variety of scattering phenomena. The prominent among them is the Raman scattering of O VI doublet $\lambda \lambda$ 1032,1038 angstrom, which often show strongly polarized features in the visible spectrum. Similar Raman scattering of Ly$\beta$ photons has also been predicted to occur in symbiotic stars, though with fewer detections and with weak polarization amplitudes. Here, we present the discovery of strong variable polarization in the H$\alpha$ profile of a recently established symbiotic system Y Gem, over a period of nearly 22 months. This is, most likely, a very rare detection of the strongly polarized Raman scattered Ly$\beta$ photons, falling at the H$\alpha$ emission. Monte-Carlo simulations have been conducted to confirm the underlying Raman scattering process causing the polarized line profile, and a simple orbital model is constructed with typical parameters available in the recent literature along with a complementary low-resolution spectroscopic data. These simulations and models are then used to validate the observed polarization variation of H$\alpha$ at different orbital phases corresponding to the epochs of observations. The possibility of such strong variable H$\alpha$ polarization, being caused by Raman scattering of Ly$\beta$, would thus open up avenues of exploring such effects in various other astrophysical situations having similar morphology.

astro-ph.SR↗

arXiv2026

To Use AI as Dice of Possibilities with Timing Computation

The dominant noun-based modeling paradigm, grounded in probability theory and committed to pre-specified noun entities as primitive modeling units, is insufficient as a \emph{grammar of thought}: It leaves \emph{timing} outside the computational scope, precluding any adequate representation of the future as an open space of possibilities. This paper addresses three conceptual gaps absent from the existing literature: (1) possibility space -- a framework admitting multiple possible timelines for the same event; (2) timing computation -- the treatment of timing as a computable rather than observed dimension; and (3) causal factum -- the maximal causal efficacy recovered by reasoning backward from possible futures, rather than assumed in advance. Together, these definitions dissolve the confounding problem inherent to noun-based causal inference and provide the foundation for a spontaneously growing causal-reasoning world model. As proof of concept, we instantiate the framework and apply it to longitudinal EHR data from 3,276 breast cancer patients, demonstrating for the first time, to our knowledge, automatic trajectory discovery and counterfactual timing deduction (i.e., a What-If Machine) in a purely data-driven manner.

cs.AI↗

arXiv2026

Artificial Intelligence Index Report 2026

Welcome to the ninth edition of the AI Index report. As AI continues to advance rapidly, the question becomes whether the systems built around it can keep up. Governance frameworks, evaluation methods, education systems, and the data infrastructure needed to track AI's impact are struggling to match the pace of the technology itself. That gap between what AI can do and how prepared we are to manage it runs through every chapter of this year's report. New in this edition, the report tracks how AI is being tested more ambitiously across reasoning, safety, and real-world task execution, and why those measurements are increasingly difficult to rely on. It also features new estimates of generative AI's economic value alongside emerging evidence of its labor market effects, an analytical framework on AI sovereignty, and a science chapter developed in collaboration with Schmidt Sciences. For the first time, the report features standalone chapters on AI in science and AI in medicine, reflecting AI's growing impact across these two domains.

cs.AI↗

arXiv2026

Development of ProtoPol: a medium resolution echelle spectro-polarimeter for PRL telescopes, Mt Abu, India -- Part I : the design, development and laboratory characterization

ProtoPol is a medium-resolution echelle spectro-polarimeter developed for Physical Research Laboratory (PRL) 1.2m and 2.5m telescopes, Mt. Abu, India. Though initially conceived to evaluate the development methodology of the echelle spectro-polarimeter, it was subsequently elevated to the level of a full-fledged back-end instrument for PRL telescopes. ProtoPol is developed on the traditional concept of using a half-wave plate with Wollaston prism to achieve the separation of two mutually orthogonal polarized beams. These separated beams are modulated and directed into an echelle spectrometer which is employs an echelle grating and two plane reflection gratings as the cross-dispersers. Therefore, the cross-dispersed spectra for two orthogonal polarized beams are recorded in multiple orders on a CCD detector. ProtoPol is designed to operate in the visible and near IR spectral range, 4000 - 9600 angstrom, with a spectral resolution ($\delta$$\lambda$) around 0.4-0.75 angstrom. The uniqueness of ProtoPol lies in its design which has entirely been developed with commercially available off-the-shelf optical and opto-mechanical components. This feature makes ProtoPol a noteworthy development as it offers a cost-effective way to develop spectro-polarimeters with such resolutions for small-aperture (2-3m) telescopes around the world, in a much shorter development period. ProtoPol has been successfully developed and commissioned on PRL 1.2m and 2.5m telescopes since December 2023, and a variety of observations have been carried out for instrument characterization, performance verification, and scientific purposes. This is the first of the two-part research articles series, wherein we present the design and development methodology of ProtoPol, along with its laboratory characterization and performance.

astro-ph.IM↗

arXiv2026

Development of ProtoPol: a medium resolution echelle spectro-polarimeter for PRL telescopes, Mt Abu, India -- Part II : the data-reduction pipeline, on-sky characterization $\&$ performance verification and first science results

We present the development of ProtoPol - a medium resolution echelle spectro-polarimeter for the PRL 1.2m and 2.5m telescopes at Mt Abu observatory, India. In this second and final part of the paper series, we report on the development of a dedicated data reduction pipeline of ProtoPol along with several characterization, performance evaluation, and scientific observations to quantify the performance of the instrument. ProtoPol provides a spectral resolution in the range of $\sim$0.4 - 0.75 angstrom across various orders in the visible wavelength range of 4000-9600 angstrom. On PRL 2.5m telescope, an SNR of 10 is achieved for $m_V\sim13.2$ source in 1 hour of integration time, and its throughput is estimated to be $\sim$6\% including all the contributing factors such as atmospheric transmission, telescope reflectivity, instrument's optics, CCD efficiency etc. ProtoPol achieved a linear polarization accuracy $\delta P \approx 0.1-0.2\%$ in 2 hours of integration time for a source with $m_V\approx8$. The instrumental polarization is determined to be around $0.1\%$. We also present the first science results with ProtoPol to demonstrate the capabilities of the instrument. A sample of Herbig Ae/Be stars, classical Herbig stars, Symbiotic stars, and AGB/post-AGB stars were observed over the period of one and half years for their spectro-polarimetry measurements covering various physical mechanisms such as intrinsic line polarization in Herbig and classical Be stars, Raman scattered features in Symbiotic stars, as well as continuum polarization in AGB/post-AGB stars to verify the polarization performance of the instrument.

astro-ph.IM↗

arXiv2026

Task Aware Modulation Using Representation Learning for Upsaling of Terrestrial Carbon Fluxes

Accurately upscaling terrestrial carbon fluxes is central to estimating the global carbon budget, yet remains challenging due to the sparse and regionally biased distribution of ground measurements. Existing data-driven upscaling products often fail to generalize beyond observed domains, leading to systematic regional biases and high predictive uncertainty. We introduce Task-Aware Modulation with Representation Learning (TAM-RL), a framework that couples spatio-temporal representation learning with knowledge-guided encoder-decoder architecture and loss function derived from the carbon balance equation. Across 150+ flux tower sites representing diverse biomes and climate regimes, TAM-RL improves predictive performance relative to existing state-of-the-art datasets, reducing RMSE by 8-9.6% and increasing explained variance (R2) from 19.4% to 43.8%, depending on the target flux. These results demonstrate that integrating physically grounded constraints with adaptive representation learning can substantially enhance the robustness and transferability of global carbon flux estimates.

cs.LG↗

arXiv2026

CarbonBench: A Global Benchmark for Upscaling of Carbon Fluxes Using Zero-Shot Learning

Accurately quantifying terrestrial carbon exchange is essential for climate policy and carbon accounting, yet models must generalize to ecosystems underrepresented in sparse eddy covariance observations. Despite this challenge being a natural instance of zero-shot spatial transfer learning for time series regression, no standardized benchmark exists to rigorously evaluate model performance across geographically distinct locations with different climate regimes and vegetation types. We introduce CarbonBench, the first benchmark for zero-shot spatial transfer in carbon flux upscaling. CarbonBench comprises over 1.3 million daily observations from 567 flux tower sites globally (2000-2024). It provides: (1) stratified evaluation protocols that explicitly test generalization across unseen vegetation types and climate regimes, separating spatial transfer from temporal autocorrelation; (2) a harmonized set of remote sensing and meteorological features to enable flexible architecture design; and (3) baselines ranging from tree-based methods to domain-generalization architectures. By bridging machine learning methodologies and Earth system science, CarbonBench aims to enable systematic comparison of transfer learning methods, serves as a testbed for regression under distribution shift, and contributes to the next-generation climate modeling efforts.

cs.LG↗

arXiv2026

Learning PDE Solvers with Physics and Data: A Unifying View of Physics-Informed Neural Networks and Neural Operators

Partial differential equations (PDEs) are central to scientific modeling. Modern workflows increasingly rely on learning-based components to support model reuse, inference, and integration across large computational processes. Despite the emergence of various physics-aware data-driven approaches, the field still lacks a unified perspective to uncover their relationships, limitations, and appropriate roles in scientific workflows. To this end, we propose a unifying perspective to place two dominant paradigms: Physics-Informed Neural Networks (PINNs) and Neural Operators (NOs), within a shared design space. We organize existing methods from three fundamental dimensions: what is learned, how physical structures are integrated into the learning process, and how the computational load is amortized across problem instances. In this way, many challenges can be best understood as consequences of these structural properties of learning PDEs. By analyzing advances through this unifying view, our survey aims to facilitate the development of reliable learning-based PDE solvers and catalyze a synthesis of physics and data.

cs.LG↗

arXiv2025

BHiCect 2.0: Multi-resolution clustering of Hi-C data

Chromatin conformation capture technologies such as Hi-C have revealed that the genome is organized in a hierarchy of structures spanning multiple scales observed at different resolutions. Current algorithms often focus on specific interaction patterns found at a specific Hi-C resolution. We present BHi-Cect 2.0, a method that leverages Hi-C data at multiple resolutions to describe chromosome architecture as nested preferentially self-interacting clusters using spectral clustering. This new version describes the hierarchical configuration of chromosomes by now integrating multiple Hi-C data resolutions. Our new implementation offers a more comprehensive description of the multi-scale architecture of the chromosomes. We further provide these functionalities as an R package to assist their integration with other computational pipelines. The BHiCect 2.0 R packages is available on github at https://github.com/princeps091-binf/BHiCect2with the version used for this manuscript on Zenodo at https://doi.org/10.5281/zenodo.17985844.

q-bio.GN↗

arXiv2024

Evaluating Deep Learning Approaches for Predictions in Unmonitored Basins with Continental-scale Stream Temperature Models

The prediction of streamflows and other environmental variables in unmonitored basins is a grand challenge in hydrology. Recent machine learning (ML) models can harness vast datasets for accurate predictions at large spatial scales. However, there are open questions regarding model design and data needed for inputs and training to improve performance. This study explores these questions while demonstrating the ability of deep learning models to make accurate stream temperature predictions in unmonitored basins across the conterminous United States. First, we compare top-down models that utilize data from a large number of basins with bottom-up methods that transfer ML models built on local sites, reflecting traditional regionalization techniques. We also evaluate an intermediary grouped modeling approach that categorizes sites based on regional co-location or similarity of catchment characteristics. Second, we evaluate trade-offs between model complexity, prediction accuracy, and applicability for more target locations by systematically removing inputs. We then examine model performance when additional training data becomes available due to reductions in input requirements. Our results suggest that top-down models significantly outperform bottom-up and grouped models. Moreover, it is possible to get acceptable accuracy by reducing both dynamic and static inputs enabling predictions for more sites with lower model complexity and computational needs. From detailed error analysis, we determined that the models are more accurate for sites primarily controlled by air temperatures compared to locations impacted by groundwater and dams. By addressing these questions, this research offers a comprehensive perspective on optimizing ML model design for accurate predictions in unmonitored regions.

cs.LG↗

arXiv2024

Hierarchical Conditional Multi-Task Learning for Streamflow Modeling

Streamflow, vital for water resource management, is governed by complex hydrological systems involving intermediate processes driven by meteorological forces. While deep learning models have achieved state-of-the-art results of streamflow prediction, their end-to-end single-task learning approach often fails to capture the causal relationships within these systems. To address this, we propose Hierarchical Conditional Multi-Task Learning (HCMTL), a hierarchical approach that jointly models soil water and snowpack processes based on their causal connections to streamflow. HCMTL utilizes task embeddings to connect network modules, enhancing flexibility and expressiveness while capturing unobserved processes beyond soil water and snowpack. It also incorporates the Conditional Mini-Batch strategy to improve long time series modeling. We compare HCMTL with five baselines on a global dataset. HCMTL's superior performance across hundreds of drainage basins over extended periods shows that integrating domain-specific causal knowledge into deep learning enhances both prediction accuracy and interpretability. This is essential for advancing our understanding of complex hydrological systems and supporting efficient water resource management to mitigate natural disasters like droughts and floods.

cs.LG↗

arXiv2024

ExoTST: Exogenous-Aware Temporal Sequence Transformer for Time Series Prediction

Accurate long-term predictions are the foundations for many machine learning applications and decision-making processes. Traditional time series approaches for prediction often focus on either autoregressive modeling, which relies solely on past observations of the target ``endogenous variables'', or forward modeling, which considers only current covariate drivers ``exogenous variables''. However, effectively integrating past endogenous and past exogenous with current exogenous variables remains a significant challenge. In this paper, we propose ExoTST, a novel transformer-based framework that effectively incorporates current exogenous variables alongside past context for improved time series prediction. To integrate exogenous information efficiently, ExoTST leverages the strengths of attention mechanisms and introduces a novel cross-temporal modality fusion module. This module enables the model to jointly learn from both past and current exogenous series, treating them as distinct modalities. By considering these series separately, ExoTST provides robustness and flexibility in handling data uncertainties that arise from the inherent distribution shift between historical and current exogenous variables. Extensive experiments on real-world carbon flux datasets and time series benchmarks demonstrate ExoTST's superior performance compared to state-of-the-art baselines, with improvements of up to 10\% in prediction accuracy. Moreover, ExoTST exhibits strong robustness against missing values and noise in exogenous drivers, maintaining consistent performance in real-world situations where these imperfections are common.

cs.LG↗

arXiv2024

Towards Knowledge Guided Pretraining Approaches for Multimodal Foundation Models: Applications in Remote Sensing

Self-supervised learning has emerged as a powerful paradigm for pretraining foundation models using large-scale data. Existing pretraining approaches predominantly rely on masked reconstruction or next-token prediction strategies, demonstrating strong performance across various downstream tasks, including geoscience applications. However, these approaches do not fully capture the knowledge of causal interplay between different geospatial and environmental variables. To address this limitation, we propose Knowledge Guided Variable-Step Forecasting (KG-VSF), a novel pretraining task that models forecasting as a conditional generation task, where driver variables (e.g., weather) inform the prediction of response variables (e.g., satellite imagery). We demonstrate that pretraining in such a fashion leads to strong embeddings which give enhanced performance when finetuned on downstream tasks where capturing this causality matters such as pixel wise crop type mapping, soil moisture estimation and forecasting, missing image prediction, and future image forecasting when compared to finetuning embeddings from other standard pretraining approaches.

cs.CV↗