455 citations · 763 across the 16 of their papers we have counts for
16 papers
Reconstructing Lyman- Fields from Low-Resolution Hydrodynamical Simulations with Deep Learning
Cooper Jacobus, Peter Harrington, Zarija Lukić
Hydrodynamical cosmological simulations are a powerful tool for accurately predicting the properties of the intergalactic medium (IGM) and for producing mock skies that can be comp…
Towards Stability of Autoregressive Neural Operators
Michael McCabe, Peter Harrington, Shashank Subramanian +1
Neural operators have proven to be a promising approach for modeling spatiotemporal systems in the physical sciences. However, training these models for large systems can be quite…
Towards Foundation Models for Scientific Machine Learning: Characterizing Scaling and Transfer Behavior
Shashank Subramanian, Peter Harrington, Kurt Keutzer +4
Pre-trained machine learning (ML) models have shown great performance for a wide range of applications, in particular in natural language processing (NLP) and computer vision (CV).…
Generative Modeling of High-resolution Global Precipitation Forecasts
James Duncan, Shashank Subramanian, Peter Harrington
Forecasting global precipitation patterns and, in particular, extreme precipitation events is of critical importance to preparing for and adapting to climate change. Making accurat…
FourCastNet: Accelerating Global High-Resolution Weather Forecasting using Adaptive Fourier Neural Operators
Thorsten Kurth, Shashank Subramanian, Peter Harrington +6
Extreme weather amplified by climate change is causing increasingly devastating impacts across the globe. The current use of physics-based numerical weather prediction (NWP) limits…
FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators
Jaideep Pathak, Shashank Subramanian, Peter Harrington +10
FourCastNet, short for Fourier Forecasting Neural Network, is a global data-driven weather forecasting model that provides accurate short to medium-range global predictions at $0.2…
Self-supervised similarity search for large scientific datasets
George Stein, Peter Harrington, Jacqueline Blaum +2
We present the use of self-supervised learning to explore and exploit large unlabeled datasets. Focusing on 42 million galaxy images from the latest data release of the Dark Energy…
Mining for Strong Gravitational Lenses with Self-supervised Learning
George Stein, Jacqueline Blaum, Peter Harrington +2
We employ self-supervised representation learning to distill information from 76 million galaxy images from the Dark Energy Spectroscopic Instrument Legacy Imaging Surveys' Data Re…
HyPhy: Deep Generative Conditional Posterior Mapping of Hydrodynamical Physics
Benjamin Horowitz, Max Dornfest, Zarija Lukić +1
Generating large volume hydrodynamical simulations for cosmological observables is a computationally demanding task necessary for next generation observations. In this work, we con…
Fast, high-fidelity Lyman forests with convolutional neural networks
Peter Harrington, Mustafa Mustafa, Max Dornfest +2
Full-physics cosmological simulations are powerful tools for studying the formation and evolution of structure in the universe but require extreme computational resources. Here, we…
Estimating Galactic Distances From Images Using Self-supervised Representation Learning
Md Abul Hayat, Peter Harrington, George Stein +2
We use a contrastive self-supervised learning framework to estimate distances to galaxies from their photometric images. We incorporate data augmentations from computer vision as w…
Self-Supervised Representation Learning for Astronomical Images
Md Abul Hayat, George Stein, Peter Harrington +2
Sky surveys are the largest data generators in astronomy, making automated tools for extracting meaningful scientific information an absolute necessity. We show that, without the n…
The Case for Strong Scaling in Deep Learning: Training Large 3D CNNs with Hybrid Parallelism
Yosuke Oyama, Naoya Maruyama, Nikoli Dryden +6
We present scalable hybrid-parallel algorithms for training large-scale 3D convolutional neural networks. Deep learning-based emerging scientific workflows often require model trai…
Enhanced mixing in magnetized fingering convection, and implications for RGB stars
Peter Harrington, Pascale Garaud
Double-diffusive convection has been well studied in geophysical contexts, but detailed investigations of the regimes characteristic of stellar or planetary interiors have only rec…
Multi-wavelength Signatures of Cosmic Rays in the Milky Way
E. Orlando, P. Harrington, A. W. Strong
Cosmic rays (CRs) propagate in the Milky Way and interact with the interstellar medium and magnetic fields. These interactions produce emissions that span the electromagnetic spect…
Partial dust obscuration in active galactic nuclei as a cause of broad-line profile and lag variability, and apparent accretion disc inhomogeneities
C. Martin Gaskell, Peter Z. Harrington
The profiles of the broad emission lines of active galactic nuclei (AGNs) and the time delays in their response to changes in the ionizing continuum ("lags") give information about…