SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2412.00568 · NeurIPS · 2024

The Well: a Large-Scale Collection of Diverse Physics Simulations for Machine Learning

Miles Cranmer, Shirley Ho, Yan-Fei Jiang, François Rozet, Bruno Régaldo-Saint Blancard, Rudy Morel, Michael Mccabe, Ruben Ohana, Liam Parker, Blakesley Burkhart, Jeff Shen, Keiya Hirashima, and 16 more

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 13 functions out of this paper's own repositories and ran 0 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
lanl/nubhlight canonical 0 of 9
PolymathicAI/the_well canonical 0 of 4
FunctionStatusWhere it lives
build_1d_power_spectrum Not yet run PolymathicAI/the_well/the_well/benchmark/metrics/plottable_data.py
code served (permissive licence) · get_code("cf1e8ae5dfb35212")
get_data Not yet run lanl/nubhlight/script/analysis/accumulate_skynet_mass_fractions.py
pointer only (licence: NOASSERTION) · get_code("81106a3d566379ad")
get_data Not yet run lanl/nubhlight/script/analysis/accumulate_skynet_output.py
pointer only (licence: NOASSERTION) · get_code("e8dede4e621a64e5")
get_env_name Not yet run lanl/nubhlight/script/config.py
pointer only (licence: NOASSERTION) · get_code("d40053a9a7e821e4")
get_files Not yet run lanl/nubhlight/script/util.py
pointer only (licence: NOASSERTION) · get_code("7990041b657e98a1")
get_header Not yet run lanl/nubhlight/script/analysis/accumulate_skynet_mass_fractions.py
pointer only (licence: NOASSERTION) · get_code("da272b91191facb4")
get_header Not yet run lanl/nubhlight/script/analysis/accumulate_skynet_output.py
pointer only (licence: NOASSERTION) · get_code("bd294fc3c02bee96")
get_metadata Not yet run lanl/nubhlight/script/analysis/accumulate_skynet_mass_fractions.py
pointer only (licence: NOASSERTION) · get_code("c6c8065f319c5d7f")
hdf5_to_xarray Not yet run PolymathicAI/the_well/the_well/utils/export.py
code served (permissive licence) · get_code("52905a530370ca29")
param_norm Not yet run PolymathicAI/the_well/the_well/benchmark/trainer/training.py
code served (permissive licence) · get_code("c43574b335efa848")
parm_is_active Not yet run lanl/nubhlight/script/util.py
pointer only (licence: NOASSERTION) · get_code("c812dff903ab1f9f")
sanitize_path Not yet run lanl/nubhlight/script/util.py
pointer only (licence: NOASSERTION) · get_code("0e652a3519220159")
vector_dtw Not yet run PolymathicAI/the_well/the_well/benchmark/metrics/temporal.py
code served (permissive licence) · get_code("eb87dbb5e169f5be")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Machine learning based surrogate models offer researchers powerful tools for accelerating simulation-based workflows. However, as standard datasets in this space often cover small classes of physical behavior, it can be difficult to evaluate the efficacy of new approaches. To address this gap, we introduce the Well: a large-scale collection of datasets containing numerical simulations of a wide variety of spatiotemporal physical systems. The Well draws from domain experts and numerical software developers to provide 15TB of data across 16 datasets covering diverse domains such as biological systems, fluid dynamics, acoustic scattering, as well as magneto-hydrodynamic simulations of extra-galactic fluids or supernova explosions. These datasets can be used individually or as part of a broader benchmark suite. To facilitate usage of the Well, we provide a unified PyTorch interface for training and evaluating models. We demonstrate the function of this library by introducing example baselines that highlight the new challenges posed by the complex dynamics of the Well. The code and data is available at https://github.com/PolymathicAI/the_well.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2412.00568")
get_code_for_paper("2412.00568")
have("2412.00568")

Connect an agent — have() is free.