Deep neural networks improve CMB lensing potential reconstruction for future cosmic microwave background experiments.
problem Low noise levels in upcoming CMB experiments require improved methods for extracting lensing potential.
method Deep convolutional neural networks (ResUNet) trained on simulated data without physical parametrization.
result ResUNets recover lensing potential with higher signal-to-noise ratio than quadratic estimator.
We demonstrate the potential of Deep Learning methods for measurements of cosmological parameters from density fields, focusing on the extraction of non-Gaussian information. We consider weak lensing mass maps as our dataset. We aim for our method to be able to distinguish between five models, which were chosen to lie …
Weak lensing maps contain information beyond two-point statistics on small scales. Much recent work has tried to extract this information through a range of different observables or via nonlinear transformations of the lensing field. Here we train and apply a 2D convolutional neural network to simulated noiseless lensi…
Machine learning helps infer dark matter substructure from strong lensing images.
problem Extracting information about dark matter substructure from strong lensing images is challenging.
method Simulation-based inference techniques and neural networks trained on simulator data.
result Efficiently trained neural networks can estimate likelihood ratios for substructure parameters.
Method generates joint posterior samples of source and foreground mass distributions for gravitational lensing.
problem Challenging inference problem for high-resolution, high signal-to-noise ratio gravitational lensing.
method Combines diffusion-based generative modeling and recurrent inference machines.
result Can model realistic gravitational lensing simulations down to the noise level.
New equations reveal how cylinder power in progressive lenses depends on geodesic curvature.
problem Current understanding of cylinder power in progressive lenses is incomplete.
method Derived complete compatibility equations for spatially-varying curvature surfaces.
result Cylinder power depends on geodesic curvature, not just principal curvature.
ResUNet-CMB neural network reconstructs CMB effects from noisy data.
problem Reconstructing CMB anisotropies from noisy data.
method Convolutional neural network (ResUNet-CMB) for simultaneous reconstruction of lensing and reionization.
result ResUNet-CMB outperforms quadratic estimators at low noise levels and avoids lensing-induced bias.
Estimates how many times a star appears due to gravitational lensing.
problem Estimating the number of times an observer sees a star due to gravitational lensing.
method Use affine linking numbers to estimate the number of times an observer sees a star.
result Estimates the number of times an observer sees a star due to gravitational lensing.
GANs create realistic galaxy images for astronomy.
problem Handling large astronomical datasets.
method Chained generative adversarial networks (GANs).
result GAN-generated galaxy images closely match real galaxies.
Enhances weak lensing inference with neural summaries.
problem Extracting additional information from weak lensing convergence maps.
method Hybrid approach combining physics-based and neural summaries.
result Neural summaries extract up to 8 times more information than angular power spectra.
Deep learning reduces noise in weak lensing mass maps using GANs.
problem Noise reduction in weak lensing mass maps.
method Generative adversarial networks (GANs) applied to Subaru Hyper Suprime-Cam data.
result GANs successfully reproduce non-Gaussian information in denoised maps, showing stronger cosmological dependence.
Paper uses ResUNet-CMB to reconstruct cosmic polarization rotation from CMB data.
problem Reconstructing anisotropic cosmic polarization rotation from CMB data.
method Extended ResUNet-CMB to handle gravitational lensing and patchy reionization.
result ResUNet-CMB outperforms standard quadratic estimator in reconstructing all three effects.
Deep learning helps remove secondary B-mode polarization to detect primordial gravitational waves.
problem Removing secondary B-mode polarization from CMB data to detect primordial gravitational waves. method Applied deep learning (ResUNet-CMB) to estimate and remove multiple sources of secondary B-mode polarization. result Deep learning can produce nearly optimal, unbiased estimates of the amplitude of primordial gravitational waves.
The modulation transfer function (MTF) is widely used to characterise the performance of optical systems. Measuring it is costly and it is thus rarely available for a given lens specimen. Instead, MTFs based on simulations or, at best, MTFs measured on other specimens of the same lens are used. Fortunately, images reco…
We provide a geometric explanation for the existence of magnification relations for the A, D, E family of caustic singularities, which were established in recent work. In particular, it was shown that for families of general mappings between planes exhibiting any of these caustic singularities, and for any non-caustic …
Machine learning detects subhalos in lensed images with high accuracy and low false positives.
problem Detecting substructure in strongly lensed images.
method Developed a neural network for image segmentation to locate and mass estimate subhalos.
result The network can detect subhalos with masses m≳108.5M⊙ and measure the subhalo mass function. The paper solves the isoperimetric problem in Riemannian optical geometry, proving circles minimize lengths with area constraints.
problem Optical geometry of static spherically symmetric spacetimes.
method Applying isoperimetric problem results to curves in Riemannian optical geometry.
result Length-minimizing curves with area constraints are circles, with implications for photon spheres.
Deep models predict gas properties from dark matter to aid cosmological simulations.
problem Computational challenges in running hydrodynamical simulations for large-scale structure and baryonic probes.
method Trained variational auto-encoders and generative adversarial networks on BAHAMAS hydrodynamical simulation data to map matter density to gas pressure.
result Generated tSZ maps are statistically consistent with those from BAHAMAS, enabling SLICS for tSZ covariance estimation.
In the Friedmann Model of the universe, cosmologists assume that spacelike slices of the universe are Riemannian manifolds of constant sectional curvature. This assumption is justified via Schur's Theorem by stating that the spacelike universe is locally isotropic. Here we define a Riemannian manifold as almost locally…
New emulator bridges simulators using conditional optimal transport.
problem Bridging simulators with minimal distortion.
method Flow-based approach to learn likelihood transport, COT-FM for optimal matching.
result Emulator accurately captures full correction between simulators.
Co-eye combines multiple symbolic representations to improve time series classification accuracy.
problem Challenges in time series classification due to domain diversity.
method Inspired by compound eyes, Co-eye uses multiple symbolic representations and hyper-parameterised lenses to classify time series data.
result Co-eye outperforms state-of-the-art techniques in accuracy and robustness across various domains.
First we review the definition of a negative point mass singularity. Then we examine the gravitational lensing effects of these singularities in isolation and with shear and convergence from continuous matter. We review the Inverse Mean Curvature Flow and use this flow to prove some new results about the mass of a sing…
We pursue a geometrical approach to gravitational lensing theory. We present a survey of the background theory of General Relativity, including particular properties of the Schwarzschild and Kerr solutions. Next we outline a proof of the Gauss Bonnet theorem and its applications to surfaces in optical geometry, as deve…
Vogt's theorem, concerning boundary angles of a convex arc with monotonic curvature (spiral arc), is taken as a starting point to establish basic properties of spirals. The theorem is expanded by removing requirements of convexity and curvature continuity; the cases of inflection and multiple windings are considered. P…
We explore the perspective of a bug living on the two-dimensional surface of a polyhedron. Images of various kinds of effects like lensing and cloaking are shown via color pictures of three viewpoints: the first person perspective of the bug, a map of the bug's viewpoint, and a look at the bug on the embedded polyhedro…
Evidence Networks simplify Bayesian model comparison for complex models.
problem Bayesian model comparison challenges with intractable likelihoods or priors.
method Loss functions and neural networks for fast, amortized estimation of Bayes factors.
result Evidence Networks provide accurate and scalable Bayes factor estimation.
New system studies trapped light paths in Euclidean space.
problem Trapping of light paths in Euclidean space with negative refractive index.
method Introduces wind-tree tiling billiards system to study trajectories of rays in Euclidean space with rectangular obstacles.
result Almost every configuration of the system traps trajectories with initial vertical direction in an infinite strip.
TopoFisher learns topological summaries by maximizing Fisher information, improving parameter efficiency and inference quality.
problem Simulation-based inference misses key information in low-order statistics, especially for non-Gaussian fields.
method TopoFisher uses a differentiable persistent-homology pipeline that learns topological summaries by maximizing local Gaussian Fisher information.
result TopoFisher recovers much of the available information and outperforms fixed topological vectorizations in weak gravitational lensing.
We give a characterization of critical points that allows us to define a metric invariant on all Riemannian manifolds M with a lower sectional curvature bound and an upper radius bound. We show there is a uniform upper volume bound for all such manifolds with an upper bound on this invariant. We generalize results by…
The paper addresses fairness in machine learning models through structural econometrics, projecting indexes into null spaces to find fair solutions.
problem Fairness concerns in machine learning, especially regarding disadvantaged groups.
method Model fairness as a linear operator, projecting indexes into null spaces to find fair solutions, balancing status quo and full fairness.
result Achieving approximate fairness by introducing a fairness penalty and balancing influences.
We propose flow-based likelihoods to accurately capture non-Gaussian data.
problem Bypassing the Gaussian assumption in scientific analyses.
method Use optimization targets of flow-based generative models to reconstruct likelihoods.
result Flow-based likelihoods can accurately capture non-Gaussian data, improving parameter constraints.
Network Lens identifies node behaviors in heterogeneous networks with high accuracy.
problem Identifying different behaviors in various parts of large heterogeneous networks.
method Zoom into network using different-sized lenses to capture local structure, weight signatures to predict node labels.
result Achieved a peak accuracy of ~42% on two networks with ~100,000 and ~1,000,000 nodes, significantly better than random.
Rubin LSST DESC uses AI/ML for dark energy research.
problem Challenges in uncertainty quantification and model robustness for AI/ML in DESC.
method Bayesian inference, physics-informed methods, validation frameworks, active learning.
result AI/ML methods are essential but require rigorous evaluation and governance.
New model reduces bias in cosmic shear measurements.
problem Bias in cosmic shear measurements due to non-well-defined ellipticity.
method Hybrid physical and deep learning Hierarchical Bayesian Model.
result Unbiased estimate of shear on realistic galaxies.
New black hole solutions with lens space horizons in 5D Kaluza-Klein theory.
problem Finding black hole solutions with specific horizon topologies.
method Formally asymptotically flat black hole solutions constructed through Kaluza-Klein reduction.
result Explicit construction of regular black hole solutions with L(p,q) horizons. Cosmic shear is a primary cosmological probe for several present and upcoming surveys investigating dark matter and dark energy, such as Euclid or WFIRST. The probe requires an extremely accurate measurement of the shapes of millions of galaxies based on imaging data. Crucially, the shear measurement must address and c…
We study the problem of asymptotically flat bi-axially symmetric stationary solutions of the vacuum Einstein equations in 5-dimensional spacetime. In this setting, the cross section of any connected component of the event horizon is a prime 3-manifold of positive Yamabe type, namely the 3-sphere S3, the ring $…
Reinterprets Granger causality with causal Bayesian networks and Reichenbach's principles.
problem Lack of a rigorous causal foundation in Granger causality.
method Reinterpreting Granger causality through Reichenbach's principles and causal Bayesian networks, implementing as c-GC.
result c-GC provides a more principled framework for causal discovery in observational datasets.
Paper studies DRO with MMD uncertainty sets and reveals connections to regularization and generalization.
problem Addressing the limitations of existing DRO uncertainty sets in machine learning.
method Introduces DRO with uncertainty sets measured via maximum mean discrepancy (MMD) and derives connections to regularization and generalization.
result Obtains an alternative proof of a generalization bound for Gaussian kernel ridge regression via DRO lens and suggests a new regularizer.
Study predicts lens performance using neural networks.
problem Predicting visual acuity from lens designs.
method Used a CNN to classify Landolt Cs and validate its ability to predict VA from induced defocus.
result Validation showed consistent offset of +0.20 logMAR from simulated VA, comparable to clinical repeatability.
Framework synthesizes programs for simulating complex models and estimating parameters.
problem Parameter estimation for complex models requires manual encoding of fixed model structures.
method Combines LLMs for program synthesis with neural simulation-based inference.
result Identifies plausible model families from open-ended prompts with high accuracy.
The study improves genus 1 bridge number bounds for satellite knots.
problem Improving bounds on genus 1 bridge number for satellite knots.
method Generalizes and proves new bounds for genus 0 and 1 bridge numbers.
result Shows b1(T)≥ωb1(K) for non-torus companion knots. By using Stationary-to-Randers correspondence (SRC), a characterization of light and time-convexity of the boundary of a region of a standard stationary (n+1)-spacetime is obtained, in terms of the convexity of the boundary of a domain in a Finsler n or (n+1)-space of Randers type. The latter convexity is analyzed in d…
GANs can bias synthetic data, affecting minority and female faces.
problem GANs can amplify biases in synthetic data augmentation.
method Examine GANs on face-shots with gender and skin tone biases.
result GANs generate biased synthetic data, skewing minority modes and features.
Introduces CSLC models to bridge deep generative models and classical algorithms.
problem Mode collapse and memorization issues in deep generative models and restrictive assumptions in classical algorithms.
method Introduces conditionally strongly log-concave (CSLC) models, factorizing data distribution into strongly log-concave conditional distributions.
result Efficient parameter estimation and sampling algorithms with theoretical guarantees for non-log-concave data distributions.
Combines BO with context to optimize binary feedback.
problem Optimizing expensive binary functions with context.
method Bayesian active learning and optimization.
result Efficiently chooses best context and parameters.
Study shows how algorithmic choices affect optimal batch sizes in neural networks.
problem Understanding how batch size impacts neural network training efficiency.
method Experiments and analysis of a simple quadratic model to study algorithmic choices.
result Preconditioned optimizers like Adam and K-FAC allow larger batch sizes before diminishing returns.
Survey of mathematical foundations for reinforcement learning.
problem Design and analysis of modern reinforcement learning algorithms.
method Organizes mathematical structures from probability, optimization, and operator theory.
result Unified mathematical entry point for researchers in various fields.