SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1904.09406 · 2019

DeepMoD: Deep learning for Model Discovery in noisy data

arXiv · PDF · Open in the Atlas

Code that ran

12 of 18 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 12 have no contract verdict; 0 failures of our instrument; 6 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/1904.09406.json · badge: /api/ran/1904.09406.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
PhIMaL/DeePyMoD canonical (classifier label)
classifier basis: census
3 ran · 2 no recorded run
PhIMaL/DeePyMoD_torch community (classifier label)
classifier basis not recorded on the link
9 ran · 4 no recorded run
FunctionStatusWhere it lives
create_deriv_data Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/utilities.py
code served (permissive licence) · get_code("d8df66bde31d0171")
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 tuple. 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

torch.manual_seed(0)

# X should be a tensor of shape (batch_size, input_dim)
# Based on the function, it accesses X.shape[0] for batch size and X.shape[1] for input dimension
X = torch.randn(2, 4)  # batch_size=2, input_dim=4

# max_order should be an integer controlling the number of derivative orders
max_order = 3

ARGS = {
    "X": X,
    "max_order": max_order,
}
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 tuple — a tuple output has no array shape of its own; none is 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
custom_board Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/output.py
code served (permissive licence) · get_code("1f16efe832654af1")
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 dict. 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

# The function takes a single parameter: number_of_terms
# which should be a list/array of integers representing the number of terms for each equation

# Looking at the function body:
# - It calls len(number_of_terms) to get number_of_eqs
# - It iterates through np.arange(number_of_eqs)
# - For each equation idx, it accesses number_of_terms[idx] and iterates through np.arange(number_of_terms[idx])
# So number_of_terms should be a list/array of positive integers

ARGS = {
    'number_of_terms': [3, 5, 2]  # 3 equations with 3, 5, and 2 terms respectively
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
l1_loss Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/losses.py
code served (permissive licence) · get_code("ce5babbac5d768e5")
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 Tensor, ndim 1, dtype family float, shape (3,) 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 torch

torch.manual_seed(0)

# Create a list of coefficient vectors (tensors)
# Each vector can have different sizes to be distinct
coeff_vector_list = [
    torch.randn(4),
    torch.randn(3),
    torch.randn(5)
]

# L1 regularization coefficient (scalar)
l1 = 0.01

ARGS = {
    "coeff_vector_list": coeff_vector_list,
    "l1": l1
}
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 Tensor, ndim 1, dtype family float, shape (3,) — shape (3,) 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
library_deriv Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/library_functions.py
code served (permissive licence) · get_code("5f59505d4fcf30ab")
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 tuple. 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 numpy as np

torch.manual_seed(0)

# Create data tensor with requires_grad=True so we can compute gradients
# Shape: (batch_size=2, features=4)
data = torch.randn(2, 4, requires_grad=True, dtype=torch.float32)

# Create a simple prediction tensor that depends on data
# We need prediction to be differentiable w.r.t. data
# Shape: (batch_size=2, 1)
prediction = (data[:, 0:1] ** 2 + data[:, 1:2]).sum(dim=1, keepdim=True)

# max_order parameter - determines how many derivative orders to compute
max_order = 2

ARGS = {
    "data": data,
    "prediction": prediction,
    "max_order": max_order,
}
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 tuple — a tuple output has no array shape of its own; none is 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
library_poly Ran
ran on a synthesized input
PhIMaL/DeePyMoD/src/deepymod/model/library.py
code served (permissive licence) · get_code("dae3cf23b6ec6f94")
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 Tensor, ndim 2, dtype family float, shape (2, 4) 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 torch
import numpy as np

torch.manual_seed(0)

# Create a small prediction tensor with shape (n_samples, 1)
# The function expects shape (n_samples x 1)
prediction = torch.randn(2, 1)

# Set max_order to a small positive integer
max_order = 3

ARGS = {
    "prediction": prediction,
    "max_order": max_order,
}
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 Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
library_poly Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/library_functions.py
code served (permissive licence) · get_code("766b51de9198371a")
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 Tensor, ndim 2, dtype family float, shape (2, 16) 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 torch
import numpy as np

torch.manual_seed(0)

prediction = torch.randn(2, 4)
max_order = 3

ARGS = {
    "prediction": prediction,
    "max_order": max_order,
}
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 Tensor, ndim 2, dtype family float, shape (2, 16) — shape (2, 16) 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
mse_loss Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/losses.py
code served (permissive licence) · get_code("f0384fcc975040e9")
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 Tensor, ndim 1, dtype family float, shape (4,) 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 torch

torch.manual_seed(0)

# Create small tensors for prediction and target
# The function computes MSE loss, so both need to have the same shape
# Batch size: 2, Feature dimension: 4
prediction = torch.randn(2, 4)
target = torch.randn(2, 4)

ARGS = {
    "prediction": prediction,
    "target": target,
}
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 Tensor, ndim 1, dtype family float, shape (4,) — shape (4,) 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
scaling_single_vec Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/sparsity.py
code served (permissive licence) · get_code("1332320a27b6f862")
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 Tensor, ndim 2, dtype family float, shape (4, 4) 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Construct arguments for scaling_single_vec
# The function expects:
# - coeff_vector: a weight vector to be scaled
# - sparse_theta: a library matrix (columns correspond to weight vector elements)
# - time_deriv: time derivative vector for scaling

# Looking at the function body:
# - scaling_time = torch.norm(time_deriv, dim=0) -> time_deriv should be 2D
# - scaling_theta = torch.norm(sparse_theta, dim=0)[:, None] -> sparse_theta should be 2D
# - coeff_vector_scaled = coeff_vector * (scaling_theta / scaling_time) -> broadcasting

# Use distinct dimensions:
# - batch/sample dimension: 2
# - feature/library dimension: 4
# - time_deriv: (2, 4) - 2 samples, 4 time points
# - sparse_theta: (2, 4) - 2 samples, 4 library terms
# - coeff_vector: (4,) or (1, 4) - one coefficient per library term

time_deriv = torch.randn(2, 4)  # (batch=2, features=4)
sparse_theta = torch.randn(2, 4)  # (batch=2, features=4)
coeff_vector = torch.randn(4)  # (features=4,)

ARGS = {
    "coeff_vector": coeff_vector,
    "sparse_theta": sparse_theta,
    "time_deriv": time_deriv,
}
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 Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) 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
string_matmul Ran
ran on a synthesized input
PhIMaL/DeePyMoD/src/deepymod/utils/utilities.py
code served (permissive licence) · get_code("a153c2b85fb576ff")
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 torch
import numpy as np
from itertools import product, combinations, chain

torch.manual_seed(0)

# Analyze the function:
# string_matmul takes list_1 and list_2 as parameters
# It uses product(list_1, list_2) to create pairs
# It concatenates elements: element[0] + element[1]
# This means list_1 and list_2 should contain strings

list_1 = ["a", "b", "c"]
list_2 = ["x", "y"]

ARGS = {
    "list_1": list_1,
    "list_2": list_2,
}
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 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
string_matmul Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/utilities.py
code served (permissive licence) · get_code("c67c418236cd21b3")
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 torch
import numpy as np
from itertools import product, combinations

torch.manual_seed(0)

# The function takes two lists and performs a "matrix multiplication" with strings
# by concatenating each element from list_1 with each element from list_2

list_1 = ["a", "b", "c"]
list_2 = ["x", "y"]

ARGS = {
    "list_1": list_1,
    "list_2": list_2,
}
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 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
terms_definition Ran
ran on a synthesized input
PhIMaL/DeePyMoD/src/deepymod/utils/utilities.py
code served (permissive licence) · get_code("d1ea38967acd7b00")
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.

# Analyze the function to understand what it needs
# terms_definition takes:
# - poly_list: a list of lists of strings
# - deriv_list: a list of lists of strings
# 
# The function:
# 1. Checks if len(poly_list) == 1
# 2. Uses string_matmul which does: [element[0] + element[1] for element in product(list_1, list_2)]
#    This concatenates strings from two lists
# 3. Uses combinations(poly_list, 2) and combinations(deriv_list, 2)
# 4. Accesses [1:] slicing on strings (skipping first character)
# 5. Uses product(poly_list, deriv_list)
#
# So we need:
# - poly_list: list of lists of strings
# - deriv_list: list of lists of strings

# Case 1: Single polynomial (len(poly_list) == 1)
poly_list_single = [['x', 'y', 'z']]
deriv_list_single = [['dx', 'dy']]

ARGS = {
    'poly_list': poly_list_single,
    'deriv_list': deriv_list_single
}
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 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
threshold_single Ran
ran on a synthesized input
PhIMaL/DeePyMoD_torch/src/deepymod_torch/sparsity.py
code served (permissive licence) · get_code("a2d26f2454752c64")
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 tuple. 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

torch.manual_seed(0)

# Create small coefficient vectors
# coeff_vector_scaled: the scaled version used for thresholding
coeff_vector_scaled = torch.randn(4)

# coeff_vector: the original unscaled version to preserve values
coeff_vector = torch.randn(4)

ARGS = {
    "coeff_vector_scaled": coeff_vector_scaled,
    "coeff_vector": coeff_vector,
}
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 tuple — a tuple output has no array shape of its own; none is 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
library_1D_in Not yet run
unverified (harvested, no recorded run)
PhIMaL/DeePyMoD_torch/src/deepymod_torch/library_functions.py
code served (permissive licence) · get_code("c5cbe10657219208")
How it ran

no invocation recorded.

library_deriv Not yet run
unverified (harvested, no recorded run)
PhIMaL/DeePyMoD/src/deepymod/model/library.py
code served (permissive licence) · get_code("cb46958576f26c1f")
How it ran

no invocation recorded.

load_tensorboard Not yet run
unverified (harvested, no recorded run)
PhIMaL/DeePyMoD/src/deepymod/analysis/load_tensorboard.py
code served (permissive licence) · get_code("d2ea40ed1296b9fe")
How it ran

no invocation recorded.

reg_loss Not yet run
unverified (harvested, no recorded run)
PhIMaL/DeePyMoD_torch/src/deepymod_torch/losses.py
code served (permissive licence) · get_code("692997db787ace60")
How it ran

no invocation recorded.

scaling Not yet run
unverified (harvested, no recorded run)
PhIMaL/DeePyMoD_torch/src/deepymod_torch/sparsity.py
code served (permissive licence) · get_code("35cc33cab5241f05")
How it ran

no invocation recorded.

terms_definition Not yet run
unverified (harvested, no recorded run)
PhIMaL/DeePyMoD_torch/src/deepymod_torch/utilities.py
code served (permissive licence) · get_code("a06431fdb99931c8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 18 samples lifted from this paper's linked repositories, 18 carry a function identity and 0 do not. Of the 16 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 16 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.

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

We introduce DeepMoD, a Deep learning based Model Discovery algorithm. DeepMoD discovers the partial differential equation underlying a spatio-temporal data set using sparse regression on a library of possible functions and their derivatives. A neural network approximates the data and constructs the function library, but it also performs the sparse regression. This construction makes it extremely robust to noise, applicable to small data sets, and, contrary to other deep learning methods, does not require a training set. We benchmark our approach on several physical problems such as the Burgers', Korteweg-de Vries and Keller-Segel equations, and find that it requires as few as $\mathcal{O}(10^2)$ samples and works at noise levels up to $75\%$. Motivated by these results, we apply DeepMoD directly on noisy experimental time-series data from a gel electrophoresis experiment and find that it discovers the advection-diffusion equation describing this system.

For agents

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

get_harvested_code_for_paper("1904.09406")
get_code_for_paper("1904.09406")
have("1904.09406")

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

curl https://syntology.ai/api/ran/1904.09406.json

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

[![Syntology run record](https://syntology.ai/api/ran/1904.09406.svg)](https://syntology.ai/paper/1904.09406)

Connect an agent — have() is free.