SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2404.11341 · 2024

The Causal Chambers: Real Physical Systems as a Testbed for AI Methodology

arXiv · PDF · Open in the Atlas

Code that ran

20 of 29 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; 1 honoured the contract we drafted, 1 violated it and 14 have no contract verdict; 4 failures of our instrument; 9 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/2404.11341.json · badge: /api/ran/2404.11341.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
juangamella/causal-chamber-paper canonical (classifier label)
classifier basis: census
6 ran · 6 no recorded run
juangamella/causal-chamber-package community (classifier label)
classifier basis not recorded on the link
14 ran · 3 no recorded run
FunctionStatusWhere it lives
bool_flag Ran
ran, violated the contract we drafted
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from cisnlp/ofa, is pointer only (licence: NONE) · get_code("cc66d68d2046d6a3")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from cisnlp/ofa, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

compute_metrics Ran
ran; our contract draft was wrong, not the code
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/metrics.py
code served (permissive licence) · get_code("69b5fa55a312675e")
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 defaultdict. 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
from collections import defaultdict

# Create sample data for the function
# The function expects infos dict with 'true' and 'predicted' keys
# Each containing lists of numpy arrays

# Create true values - list of 1D numpy arrays
true_values = [
    np.array([1.0, 2.0, 3.0, 4.0]),
    np.array([5.0, 6.0, 7.0, 8.0]),
]

# Create predicted values - list of 1D numpy arrays
predicted_values = [
    np.array([1.1, 2.1, 2.9, 4.1]),
    np.array([5.2, 5.9, 7.1, 7.9]),
]

# Create infos dictionary
infos = {
    'true': true_values,
    'predicted': predicted_values,
}

# Define metrics to compute
metrics = 'r2,r2_zero,accuracy_l1,_mse,_nmse,_rmse'

ARGS = {
    'infos': infos,
    'metrics': metrics,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type defaultdict — a defaultdict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a defaultdict 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
coord_grid Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/image_capture.py
code served (permissive licence) · get_code("64e1fb55950c1610")
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 ndarray, ndim 3, dtype family int, shape (4, 4, 2) 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 numpy as np

def coord_grid(image_size):
    X = np.tile(np.arange(image_size), (image_size, 1))
    Y = X.T[::-1, :]
    grid = np.array([X, Y])
    return np.transpose(grid, (1, 2, 0))

ARGS = {
    "image_size": 4
}
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 ndarray, ndim 3, dtype family int, shape (4, 4, 2) — shape (4, 4, 2) 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
create_logger Ran
ran; our contract draft was wrong, not the code
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/logger.py
pointer only (licence: MIT); the stored copy get_code serves, from rahul3/LAWT, is pointer only (licence: NOASSERTION) · get_code("2f7db44ecae76417")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from rahul3/LAWT, and GitHub could not classify the licence of that repository (NOASSERTION). The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

dHSIC_calc Ran
ran, honoured the contract we drafted
juangamella/causal-chamber-paper/case_studies/src/hsic.py
pointer only (licence: MIT); the stored copy get_code serves, from Alex-Markham/repare, is pointer only (licence: GPL-3.0) · get_code("ea1ee09f34621131")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Alex-Markham/repare, and GPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

discrete_rw Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/datasets/sampling.py
code served (permissive licence) · get_code("40e3dc59408a02c9")
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 ndarray, ndim 1, dtype family float, shape (10,) 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 numpy as np

ARGS = {
    "n": 10,
    "minn": 0,
    "maxx": 100,
    "step_sizes": [-1, 0, 1],
    "p": None,
    "seed": 42
}
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 ndarray, ndim 1, dtype family float, shape (10,) — shape (10,) 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
edges_to_adjacency Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/utils.py
code served (permissive licence) · get_code("4188cf779343ba4b")
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 ndarray, 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 numpy as np

# Define variables - these are node identifiers in a graph
variables = ['A', 'B', 'C', 'D']

# Define edges as a list of tuples (from, to)
edges = [
    ('A', 'B'),
    ('B', 'C'),
    ('A', 'C'),
    ('C', 'D'),
]

ARGS = {
    'edges': edges,
    'variables': variables,
}
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 ndarray, 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
gaussian_grammat Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
juangamella/causal-chamber-paper/case_studies/src/hsic.py
pointer only (licence: MIT); the stored copy get_code serves, from Alex-Markham/repare, is pointer only (licence: GPL-3.0) · get_code("df07fe3baf2d9722")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Alex-Markham/repare, and GPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

graph_to_tikz Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/utils.py
code served (permissive licence) · get_code("3006ede7a0ce2123")
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 str. 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

# Create a small adjacency matrix for testing
# 4x4 matrix with some non-zero entries to represent edges
A = np.array([
    [0, 1, 0, 1],
    [0, 0, 1, 0],
    [1, 0, 0, 1],
    [0, 1, 0, 0]
], dtype=float)

# Generate points on circle for 4 nodes
points = [
    (0.0, 1.0),
    (1.0, 0.0),
    (0.0, -1.0),
    (-1.0, 0.0)
]

# Custom labels
labels = ['$A$', '$B$', '$C$', '$D$']

ARGS = {
    'A': A,
    'points': points,
    'radius': 1.5,
    'labels': labels,
    'spaces': 4,
    'bend': None,
    'undirected': False
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
hexagon_mask Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/image_capture.py
code served (permissive licence) · get_code("9709404544a135d2")
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 ndarray, ndim 2, dtype family float, shape (64, 64) 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 numpy as np

# Define arguments for hexagon_mask function
ARGS = {
    "center_x": 0.5,
    "center_y": 0.5,
    "radius": 0.3,
    "offset": 0.0,
    "image_size": 64
}
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 ndarray, ndim 2, dtype family float, shape (64, 64) — shape (64, 64) 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
hexagon_vertices Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/image_capture.py
code served (permissive licence) · get_code("a7e01aac39222e52")
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 ndarray, ndim 2, dtype family float, shape (6, 2) 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 numpy as np

# hexagon_vertices takes 4 parameters: center_x, center_y, radius, offset
# All are numeric values used in trigonometric calculations
# center_x, center_y: float coordinates for the center of the hexagon
# radius: float for the distance from center to vertices
# offset: float for angular offset in radians

ARGS = {
    "center_x": 5.0,
    "center_y": 3.0,
    "radius": 2.5,
    "offset": np.pi / 6
}
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 ndarray, ndim 2, dtype family float, shape (6, 2) — shape (6, 2) 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
model_a1 Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/wind_tunnel_models.py
code served (permissive licence) · get_code("c7f1c1cf79cd9544")
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 ndarray, 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 numpy as np

# Analyze the function:
# - L: used with np.atleast_1d(), np.maximum(), indexing with L == 0
#   Should be a numeric array or scalar
# - L_min: used in np.maximum(L, L_min), should be numeric scalar
# - omega_max: used in multiplication, should be numeric scalar

# Create test arguments
L_values = np.array([0.0, 0.5, 1.0, 1.5])
L_min_value = 0.3
omega_max_value = 100.0

ARGS = {
    "L": L_values,
    "L_min": L_min_value,
    "omega_max": omega_max_value
}
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 ndarray, 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
model_a2 Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/wind_tunnel_models.py
code served (permissive licence) · get_code("3a5d174bd964aec1")
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 ndarray, ndim 1, dtype family float, shape (5,) 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 numpy as np

# Define a callable tau function that takes loads and returns torques
def tau_func(loads):
    """Convert loads to torques"""
    return 0.5 * loads + 0.1

# Create sample data
np.random.seed(0)
loads = np.array([1.0, 1.5, 2.0, 2.5, 3.0], dtype=float)
timestamps = np.array([0.0, 0.1, 0.2, 0.3, 0.4], dtype=float)

# Define parameters
I = 0.05  # moment of inertia
K = 0.02  # friction coefficient
omega_0 = 10.0  # initial angular velocity

ARGS = {
    'loads': loads,
    'I': I,
    'tau': tau_func,
    'K': K,
    'omega_0': omega_0,
    'timestamps': timestamps,
    'simulation_steps': 100,
    'method': 'euler'
}
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 ndarray, ndim 1, dtype family float, shape (5,) — shape (5,) 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
model_b1 Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/wind_tunnel_models.py
code served (permissive licence) · get_code("29316726c2ebb066")
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 ndarray, 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 numpy as np

# Construct arguments for model_b1
# The function takes L (load values), C_min, C_max, and L_min (threshold)
# L can be a scalar or array-like
# C_min, C_max, L_min are scalar parameters

ARGS = {
    "L": np.array([0.0, 0.5, 1.0, 2.0]),  # Array of load values
    "C_min": 0.1,  # Minimum current
    "C_max": 1.0,  # Maximum current
    "L_min": 0.3,  # Load threshold
}
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 ndarray, 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
model_e1 Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/models/light_tunnel_models.py
code served (permissive licence) · get_code("b36a092e478f216d")
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 ndarray, 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 numpy as np

# Analyze the function signature and body
# model_e1(theta_1, theta_2, I0, Tp, Tc)
# - theta_1: used in np.cos(theta_1 - theta_2) ** 2, so numeric (scalar or array)
# - theta_2: used in np.cos(theta_1 - theta_2) ** 2, so numeric (scalar or array)
# - I0: multiplied with malus_factor, so numeric (scalar or array)
# - Tp: subtracted from Tc and used in multiplication, so numeric (scalar)
# - Tc: subtracted from Tp and added, so numeric (scalar)

# Create small numeric values for all parameters
theta_1 = np.array([0.1, 0.2, 0.3, 0.4])  # 1D array, 4 elements
theta_2 = np.array([0.05, 0.15, 0.25, 0.35])  # 1D array, 4 elements
I0 = np.array([1.0, 1.5, 2.0, 2.5])  # 1D array, 4 elements
Tp = 100.0  # scalar
Tc = 50.0  # scalar

ARGS = {
    "theta_1": theta_1,
    "theta_2": theta_2,
    "I0": I0,
    "Tp": Tp,
    "Tc": Tc,
}
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 ndarray, 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
points_on_circle Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/utils.py
code served (permissive licence) · get_code("88810e71bfca6f77")
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 2, 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

ARGS = {
    "n": 5,
    "radius": 1.0,
}
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 list, ndim 2, 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
random_walk Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/datasets/sampling.py
code served (permissive licence) · get_code("1473743688620979")
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 ndarray, ndim 1, dtype family int, shape (10,) 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 numpy as np

ARGS = {
    "n": 10,
    "step_size": 2,
    "step_resolution": 1,
    "random_state": 42,
}
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 ndarray, ndim 1, dtype family int, shape (10,) — shape (10,) 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
random_walk_mirrored Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/datasets/sampling.py
code served (permissive licence) · get_code("6d122ddfa2b80194")
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 ndarray, ndim 1, dtype family int, shape (10,) 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 numpy as np

# Construct arguments for random_walk_mirrored
# Parameters: n, minn, maxx, step_size=1, step_resolution=1, random_state=None

ARGS = {
    "n": 10,  # number of steps in the random walk
    "minn": 0,  # minimum boundary for mirroring
    "maxx": 100,  # maximum boundary for mirroring
    "step_size": 2,  # size of each step
    "step_resolution": 1,  # resolution of steps
    "random_state": 42,  # seed for reproducibility
}
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 ndarray, ndim 1, dtype family int, shape (10,) — shape (10,) 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
set_and_measure Ran
ran on a synthesized input
juangamella/causal-chamber-package/causalchamber/datasets/utils.py
code served (permissive licence) · get_code("54486aee2089156e")
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 numpy as np

ARGS = {
    "value_ranges": {
        "temperature": {20, 21, 22, 23, 24, 25},
        "pressure": {1.0, 1.1, 1.2, 1.3},
        "voltage": {5, 6, 7, 8, 9}
    },
    "n_interventions": 3,
    "n_msr": 10,
    "repeat": False,
    "wait_before_set": (100, 200),
    "wait_before_msr": (50, 150),
    "random_state": 42
}
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 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
timeout Ran
ran; our contract draft was wrong, not the code
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/utils.py
code served (permissive licence) · get_code("2fe9d54ddd8520af")
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 function. 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 os
import errno
import signal
from functools import partial

# The timeout function is a decorator factory that takes parameters
# Looking at the function signature: timeout(seconds=10, error_message=os.strerror(errno.ETIME))

# Parameters:
# - seconds: int, number of seconds for timeout (default 10)
# - error_message: str, error message to use (default os.strerror(errno.ETIME))

ARGS = {
    'seconds': 5,
    'error_message': 'Operation timed out'
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
centering Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-paper/case_studies/src/hsic.py
pointer only (licence: MIT); the stored copy get_code serves, from Alex-Markham/repare, is pointer only (licence: GPL-3.0) · get_code("860aa91e378e0710")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from Alex-Markham/repare, and GPL-3.0, as GitHub detects it on that repository, is not in the permissive set Syntology serves inline. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. No invocation recorded.

get_infinite_relative_error Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/regressors.py
code served (permissive licence) · get_code("01cbe9c6ae58100c")
How it ran

no invocation recorded.

get_optimizer Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/optim.py
code served (permissive licence) · get_code("9290aab8c894b039")
How it ran

no invocation recorded.

initialize_exp Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/utils.py
code served (permissive licence) · get_code("4b15c33783dc239e")
How it ran

no invocation recorded.

order_data Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/regressors.py
code served (permissive licence) · get_code("91a6b94673648495")
How it ran

no invocation recorded.

select_dico_index Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-paper/case_studies/src/symbolicregression/envs/environment.py
code served (permissive licence) · get_code("ac7f89e9241e6a4d")
How it ran

no invocation recorded.

simulator_a1_c2 Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-package/causalchamber/models/wind_tunnel_simulators.py
code served (permissive licence) · get_code("84d3204b8e19e60c")
How it ran

no invocation recorded.

simulator_a1_c3 Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-package/causalchamber/models/wind_tunnel_simulators.py
code served (permissive licence) · get_code("158be9e8feb66308")
How it ran

no invocation recorded.

simulator_a2_c3 Not yet run
unverified (harvested, no recorded run)
juangamella/causal-chamber-package/causalchamber/models/wind_tunnel_simulators.py
code served (permissive licence) · get_code("26c8655502526376")
How it ran

no invocation recorded.

Also present in other papers' code

Of 29 samples lifted from this paper's linked repositories, 29 carry a function identity and 0 do not. Of the 29 distinct functions with an identity, 8 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 5 later) and 21 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
bool_flag 5 other papers
3 dated earlier, 2 later
1911.02116 (2019-11)
Unsupervised Cross-lingual Representation Learning at Scale
2110.03501 (2021-10)
Pretrained Language Models are Symbolic Mathematics Solvers too!
2311.08849 (2023-11)
OFA: A Framework of Initializing Unseen Subword Embeddings for Efficient Large-scale Multilingual Continued Pretraining
and 2 more not named here
create_logger 4 other papers
2 dated earlier, 2 later
2110.03501 (2021-10)
Pretrained Language Models are Symbolic Mathematics Solvers too!
2210.03403 (2022-10)
TAN Without a Burn: Scaling Laws of DP-SGD
2411.00566 (2024-11)
PatternBoost: Constructions in Mathematics with a Little Help from AI
and 1 more not named here
centering 1 other paper
1 later
2601.10531 (2026-01)
Coarsening Causal DAG Models
dHSIC_calc 1 other paper
1 later
2601.10531 (2026-01)
Coarsening Causal DAG Models
gaussian_grammat 1 other paper
1 later
2601.10531 (2026-01)
Coarsening Causal DAG Models
get_infinite_relative_error 1 other paper
1 dated earlier
2310.05573 (2023-10)
ODEFormer: Symbolic Regression of Dynamical Systems with Transformers
get_optimizer 1 other paper
1 dated earlier
2310.05573 (2023-10)
ODEFormer: Symbolic Regression of Dynamical Systems with Transformers
order_data 1 other paper
1 dated earlier
2310.05573 (2023-10)
ODEFormer: Symbolic Regression of Dynamical Systems with Transformers

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

In some fields of AI, machine learning and statistics, the validation of new methods and algorithms is often hindered by the scarcity of suitable real-world datasets. Researchers must often turn to simulated data, which yields limited information about the applicability of the proposed methods to real problems. As a step forward, we have constructed two devices that allow us to quickly and inexpensively produce large datasets from non-trivial but well-understood physical systems. The devices, which we call causal chambers, are computer-controlled laboratories that allow us to manipulate and measure an array of variables from these physical systems, providing a rich testbed for algorithms from a variety of fields. We illustrate potential applications through a series of case studies in fields such as causal discovery, out-of-distribution generalization, change point detection, independent component analysis, and symbolic regression. For applications to causal inference, the chambers allow us to carefully perform interventions. We also provide and empirically validate a causal model of each chamber, which can be used as ground truth for different tasks. All hardware and software is made open source, and the datasets are publicly available at causalchamber.org or through the Python package causalchamber.

For agents

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

get_harvested_code_for_paper("2404.11341")
get_code_for_paper("2404.11341")
have("2404.11341")

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

curl https://syntology.ai/api/ran/2404.11341.json

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

[![Syntology run record](https://syntology.ai/api/ran/2404.11341.svg)](https://syntology.ai/paper/2404.11341)

Connect an agent — have() is free.