Light curves describe temporal variations in the brightness of celestial objects. Learning robust representations of light curves is essential for large-scale automatic discovery in the dynamic universe, but existing time-series foundation models often struggle with the uneven sampling, complex noise, and wide range of physical timescales that characterize astronomical observations. We propose a domain-informed representation learning framework for irregular astronomical time series with Joint-Embedding predictive architecture (JEPA), combining semantics-preserving views, uncertainty-aware tokenization, and multi-view self-distillation. The encoders are trained with multi-view self-distillation using LeJEPA regularization on the LEAVES dataset and evaluated on the StarEmbed classification benchmark. On StarEmbed, our model outperforms hand-crafted features on 15 of 16 classification metrics. In few-shot linear probing, it achieves macro-F1 scores of 42.56 $\pm$ 7.21 with one sample per class and 63.58 $\pm$ 1.20 with 100 samples per class, consistently improving over hand-crafted features. Beyond variable-star classification, the learned representation supports similarity search, parameter estimation, and photometric zero-point drift detection. We further evaluate cross-domain adaptation on 12 heterogeneous irregular time-series datasets from PYRREGULAR, where the adapted variant matches or exceeds previous state-of-the-art performance on 5 datasets, compared with at most 3 wins by any single prior baseline. These results demonstrate that domain-informed multi-view self-distillation is an effective strategy for learning representations of irregular time series, while also highlighting that successful time-series representation learning requires domain-specific inductive biases rather than a universally optimal architecture.
We introduce a method for visual and auditory feedback when exploring the fit of a model to data. Starting with a best-fit curve fit to data, the user can drag the curve to a new position and the computer will emit a squeal, becoming louder and more unpleasant as the discrepancy between curve and data increases. We demonstrate with four examples: a two-parameter curve fit to golf putting data, a four-parameter curve fit to dilution assays, a fit to cosmological data sensitive to the parameters of the Big Bang model, and a nonparametric Gaussian process fit to temperature readings.
We investigate the reconstruction of holographic duals for strongly coupled quantum field theories in regimes characterized by large hierarchies and the presence of false vacua. Within the gauge/gravity duality, these features translate into non-trivial thermodynamic behaviour and exotic renormalization group flows, including skipping flows between non-adjacent fixed points. Building on previous work based on Physics-Informed Neural Networks (PINNs), we extend the holographic inverse problem of reconstructing the bulk scalar potential from boundary thermodynamic data into this new regime. This setting presents a variety of conceptual and numerical challenges, such as near-degenerate states, large hierarchies of energy scales, and regions of the potential that are not directly probed by the input data. We develop a set of methodological advances that overcome these obstacles, thereby improving the established PINNs-based methodology and extending it to new physical regimes of interest that were previously out of reach. Applying the developed framework, we demonstrate accurate reconstruction of scalar potentials deep into the false vacuum regime, achieving robust agreement with the physical features of the underlying thermodynamics despite significant numerical stiffness. Our results extend the bridge between holography and machine learning, and suggest that data-driven approaches can provide new insights into the structure of strongly coupled systems.
Continuous physical fields represent a large fraction of data under scientific investigation. Their multiscale structures are central to discovery, yet useful coordinates are not known in advance. Standard self-supervised methods define context and targets in fixed image coordinates, posing a predictive task misaligned with fields organized across a continuous scale hierarchy. We introduce ScaleAware-JEPA, a framework that constructs dense, label-free latent coordinates for continuous scalar fields. Constrained Diffusion Decomposition (CDD) separates each field into pixel-registered scale components and provides the scale coordinates that define the masking geometry. The resulting JEPA objective predicts hidden structure with a context footprint tied to the diffusion scale of each component rather than to an arbitrary patch size. Across MHD turbulence, interstellar molecular gas and urban nighttime-light structure, the learned geometry maps back to coherent morphology, forming dense structural atlases without labels or predefined segmentation rules. By tying latent prediction to the scale hierarchy of a field, ScaleAware-JEPA constructs latent coordinates through which complex physical patterns can be inspected before their relevant structures have been prescribed. Code is available at https://github.com/gxli/SA-JEPA.
Generative Astrodynamics is advanced in this work by extending generative modelling to an orbit determination problem in the cislunar environment. The task is formulated as conditional density estimation, aiming to infer the probability distribution of the initial state from angles-only measurements over short observation arcs. A normalising flow is trained on perturbed topocentric observations from Near Rectilinear Halo Orbits, enabling a flexible and potentially multimodal posterior representation. Given new measurements, the learned density is sampled to generate statistically consistent and physics-informed state hypotheses. These estimates are refined via nonlinear least-squares minimisation, providing a competitive warm start for classical algorithms.
We investigate the multipolar surface magnetic-field structure of the high-mass millisecond pulsar PSR J0740+6620 using the 32-bin bolometric NICER pulse profile of Dittmann et al. (2024). Building on the neural-network surrogate framework of Olmschenk et al. (2025), we model the emitting regions as open-field-line footpoints of an offset dipole plus axisymmetric quadrupole static vacuum field, rather than as prescribed geometric hotspots. We fix the stellar mass, radius, observer inclination, and hotspot temperature ratio to the Dittmann et al. (2024) maximum-likelihood values and explore the resulting 11-dimensional magnetic-field space. To make this feasible, we train convolutional neural-network surrogates on $5.12\times10^7$ synthetic bolometric light curves and use them in a parallel ensemble Markov Chain Monte Carlo calculation on 4000 CPU cores, accelerating likelihood evaluations by a factor of $\gtrsim 400$. We perform independent inferences for two calibrated temperature-weight prescriptions, Tw=1.31 and Tw=1.41, encoding the relative bolometric weight associated with the hotspot temperature difference. The posteriors, posterior-predictive light curves, and maximum-likelihood values are very similar, indicating weak sensitivity to this choice. The offset model reproduces the observed double-peaked profile and yields broad, multimodal posteriors, reflecting both the background-dominated data and degeneracies of the multipolar parameterization. The hotspot-density map shows that pulse phases constrain the approximate azimuthal placement of the emission, while latitude, surface extent, and morphology remain weakly constrained. A restricted zero offset run is disfavored within the adopted field basis. This work extends neural-network-accelerated magnetic-field inference to PSR J0740+6620 and motivates future energy-dependent, force-free, and joint X-ray/$γ$-ray extensions.