SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.05237 · 2021

Physics-based Deep Learning

arXiv · PDF · Open in the Atlas

Code that ran

7 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 0 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2109.05237.json · badge: /api/ran/2109.05237.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
thunil/Deep-Flow-Prediction canonical (classifier label)
classifier basis: census
4 ran · 0 no recorded run
tum-pbs/pbdl-dataset canonical (classifier label)
classifier basis: census
3 ran · 0 no recorded run
FunctionStatusWhere it lives
LoaderNormalizer Ran
ran on a synthesized input
thunil/Deep-Flow-Prediction/train/dataset.py
code served (permissive licence) · get_code("45f93ade2a179e74")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned TurbDataset. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import tempfile
import os

class TurbDataset:
    def __init__(self, dataDir):
        self.dataDir = dataDir
        self.totalLength = 0
        self.inputs = None
        self.targets = None

# Create a temporary directory with mock data files
temp_dir = tempfile.mkdtemp()

# Create mock .npz files
for i in range(3):
    # Create data with shape (6, 128, 128): 3 inputs + 3 targets
    data = np.random.randn(6, 128, 128).astype(np.float32)
    file_path = os.path.join(temp_dir, f'data_{i}.npz')
    np.savez(file_path, a=data)

# Create data object
data = TurbDataset(temp_dir + '/')

ARGS = {
    'data': data,
    'isTest': False,
    'shuffle': 0,
    'dataProp': None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type TurbDataset — a TurbDataset output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a TurbDataset output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
blockUNet Ran
ran on a synthesized input
thunil/Deep-Flow-Prediction/train/DfpNet.py
code served (permissive licence) · get_code("77dc7eba58f81d1f")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Sequential. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'in_c': 3,
    'out_c': 64,
    'name': 'block1',
    'transposed': False,
    'bn': True,
    'relu': True,
    'size': 4,
    'pad': 1,
    'dropout': 0.5,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Sequential — a Sequential output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Sequential output the recorded ndim 1 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
computeLR Ran
ran on a synthesized input
thunil/Deep-Flow-Prediction/train/utils.py
code served (permissive licence) · get_code("f527235b2c5581e6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import math

ARGS = {
    'i': 10,
    'epochs': 100,
    'minLR': 0.0001,
    'maxLR': 0.1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
find_absmax Ran
ran on a synthesized input
thunil/Deep-Flow-Prediction/train/dataset.py
code served (permissive licence) · get_code("dc27f11e4f37e00c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float64, ndim 0, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import torch

class MockData:
    def __init__(self, inputs, targets, total_length):
        self.inputs = inputs
        self.targets = targets
        self.totalLength = total_length

# Create small tensors for inputs and targets
# inputs and targets are lists where each element can be indexed with [x]
torch.manual_seed(0)

# Create 2 samples, each with a tensor of shape (4, 3)
# This allows indexing with [x] where x is an index into the second dimension
inputs = [
    np.random.randn(4, 3),
    np.random.randn(4, 3)
]

targets = [
    np.random.randn(4, 3),
    np.random.randn(4, 3)
]

data = MockData(inputs=inputs, targets=targets, total_length=2)

ARGS = {
    "data": data,
    "use_targets": 0,
    "x": 1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_const_sim Ran
ran on a synthesized input
tum-pbs/pbdl-dataset/pbdl/utilities.py
code served (permissive licence) · get_code("eb7fad14e3162b5b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import h5py
import numpy as np
import tempfile
import os

# Create a temporary HDF5 file with the required structure
temp_dir = tempfile.mkdtemp()
hdf5_path = os.path.join(temp_dir, 'test_data.h5')

with h5py.File(hdf5_path, 'w') as f:
    # Create the sims group
    sims_group = f.create_group('sims')
    
    # Set Constants attribute on the sims group
    constants_list = ['const1', 'const2', 'const3']
    sims_group.attrs['Constants'] = constants_list
    
    # Create a specific sim dataset (sim0) with the constants as attributes
    sim_dataset = sims_group.create_dataset('sim0', data=[1, 2, 3])
    sim_dataset.attrs['const1'] = 10.5
    sim_dataset.attrs['const2'] = 20.5
    sim_dataset.attrs['const3'] = 30.5

# Open the file and create dset
dset = h5py.File(hdf5_path, 'r')

ARGS = {
    'dset': dset,
    'sim': 0
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
h5py, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_sel_const_sim Ran
ran on a synthesized input
tum-pbs/pbdl-dataset/pbdl/utilities.py
code served (permissive licence) · get_code("85fd159e1b36b14a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from collections import defaultdict

# Create a minimal mock HDF5-like dataset structure
class MockHDF5Group:
    def __init__(self):
        self.attrs = {}
        self._children = {}
    
    def __getitem__(self, key):
        if key not in self._children:
            self._children[key] = MockHDF5Group()
        return self._children[key]

# Construct the mock dataset
dset = MockHDF5Group()

# Set up the structure: dset['sims/sim0'].attrs with some keys
dset['sims/sim0'].attrs = {
    'key1': 10,
    'key2': 20,
    'key3': 30,
    'key4': 40
}

# Set up dset['sims/'].attrs with Constants list
dset['sims/'].attrs = {
    'Constants': ['key1', 'key3']
}

# Parameters for the function call
sim = 0  # sim parameter - will be converted to string in function
sel_const = None  # None will trigger use of default Constants from dset['sims/'].attrs
# Alternative: sel_const = ['key2', 'key4']  # If we want to override with selected constants

ARGS = {
    'dset': dset,
    'sim': sim,
    'sel_const': sel_const
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_sel_const_sim_v2 Ran
ran on a synthesized input
tum-pbs/pbdl-dataset/pbdl/utilities.py
code served (permissive licence) · get_code("77330e18ab3ef26a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import h5py
import tempfile
import os

# Create a temporary HDF5 file to serve as the mock dataset
temp_dir = tempfile.mkdtemp()
hdf5_path = os.path.join(temp_dir, "mock_dataset.h5")

with h5py.File(hdf5_path, "w") as f:
    # Create the sims group
    sims_group = f.create_group("sims")
    
    # Add default Constants to sims group attrs
    sims_group.attrs["Constants"] = ["const1", "const2"]
    
    # Create sim0 group
    sim0_group = sims_group.create_group("sim0")
    
    # Create dataset 0 in sim0
    dataset_0 = sim0_group.create_dataset("0", data=[1, 2, 3])
    
    # Add attributes to dataset 0
    dataset_0.attrs["const1"] = 10
    dataset_0.attrs["const2"] = 20
    dataset_0.attrs["const3"] = 30

# Open the HDF5 file as the dset
dset = h5py.File(hdf5_path, "r")

ARGS = {
    "dset": dset,
    "sim": 0,
    "sel_const": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
h5py — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 0 later) and 5 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
computeLR 1 other paper
1 dated earlier
1810.08217 (2018-10)
Deep Learning Methods for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flows
find_absmax 1 other paper
1 dated earlier
1810.08217 (2018-10)
Deep Learning Methods for Reynolds-Averaged Navier-Stokes Simulations of Airfoil Flows

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

This document is a hands-on, comprehensive guide to deep learning in the realm of physical simulations. Rather than just theory, we emphasize practical application: every concept is paired with interactive Jupyter notebooks to get you up and running quickly. Beyond traditional supervised learning, we dive into physical loss-constraints, differentiable simulations, diffusion-based approaches for probabilistic generative AI, as well as reinforcement learning and advanced neural network architectures. These foundations are paving the way for the next generation of scientific foundation models. We are living in an era of rapid transformation. These methods have the potential to redefine what's possible in computational science.

For agents

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

get_harvested_code_for_paper("2109.05237")
get_code_for_paper("2109.05237")
have("2109.05237")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2109.05237.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2109.05237.svg)](https://syntology.ai/paper/2109.05237)

Connect an agent — have() is free.