SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2306.13802 · NeurIPS · 2023

Using topological data analysis to compare inter-subject variability across resting state functional MRI brain representations

Ty Easley, Kevin Freese, Elizabeth Munch, Janine Bijsterbosch

arXiv · PDF · Open in the Atlas

Code that ran

18 of 26 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, 0 violated it and 14 have no contract verdict; 3 failures of our instrument; 8 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/2306.13802.json · badge: /api/ran/2306.13802.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
tyo8/interval-matching-precomp_metric canonical (classifier label)
classifier basis: census
14 ran · 7 no recorded run
tyo8/brain_representations canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
tyo8/interval-matching_bootstrap extension (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
FunctionStatusWhere it lives
Jaccard Ran
ran, honoured the contract we drafted
tyo8/interval-matching_bootstrap/utils_match/compute.py
code served (permissive licence) · get_code("cfeb82aad91ec402")
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) 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 torch
import numpy as np

torch.manual_seed(0)

# Analyzing the function:
# - b1, d1, b2, d2 are used with max(), min(), and arithmetic operations
# - They appear to be numeric values (likely floats or ints)
# - The function computes a Jaccard index between two intervals [b1, d1] and [b2, d2]

# Setting up simple numeric arguments for the Jaccard function
b1 = 1.0  # beginning of first interval
d1 = 5.0  # end of first interval
b2 = 3.0  # beginning of second interval
d2 = 7.0  # end of second interval

ARGS = {
    "b1": b1,
    "d1": d1,
    "b2": b2,
    "d2": d2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, torch — 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
affine_embed_ptcloud Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/visualization/toy_models.py
code served (permissive licence) · get_code("4aa9032292690efb")
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 (7, 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

# ptcloud should be a numpy array with shape (source_dim, num_points)
# embed_dim should be an integer greater than source_dim
ptcloud = np.random.randn(3, 4)  # 3D point cloud with 4 points
embed_dim = 7

ARGS = {
    "ptcloud": ptcloud,
    "embed_dim": embed_dim,
}
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 (7, 4) — shape (7, 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
collate_simplist_data Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/collate_tagged_data.py
code served (permissive licence) · get_code("ff4ffe6b4b12e895")
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 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 os
import tempfile
import ast

# Create a temporary directory for test files
tmpdir = tempfile.mkdtemp()

# Create some test files with valid Python list data
test_file_1 = os.path.join(tmpdir, "test_data_1.txt")
test_file_2 = os.path.join(tmpdir, "test_data_2.txt")

with open(test_file_1, 'w') as f:
    f.write("[[1, 2, 3], [4, 5, 6]]")

with open(test_file_2, 'w') as f:
    f.write("[[7, 8, 9]]")

# Create the filelist argument
filelist = [test_file_1, test_file_2]

ARGS = {
    "filelist": filelist
}
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 2, 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
find_occurences_list Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/matching.py
code served (permissive licence) · get_code("d30ab60bbf7b0529")
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

torch.manual_seed(0)

ARGS = {
    "values": [1, 2, 3, 2, 4, 2, 5],
    "val": 2
}
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 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
generate_famidx_tag Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/generate_subindex.py
code served (permissive licence) · get_code("bef8c9ea8810df83")
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

np.random.seed(42)

# Create test data for generate_famidx_tag
# fam_struct: array indicating family ID for each subject
fam_struct = np.array([1, 1, 2, 2, 3, 3, 4, 4, 5])

# fam_unq: unique family IDs
fam_unq = np.array([1, 2, 3, 4, 5])

# fam_occ: occurrence count for each family ID
fam_occ = np.array([2, 2, 2, 2, 1])

ARGS = {
    "fam_struct": fam_struct,
    "fam_unq": fam_unq,
    "fam_occ": fam_occ,
    "sample_prop": 0.8,
    "verbose": 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
generate_subidx_tag Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/generate_subindex.py
code served (permissive licence) · get_code("e906073d0967a88b")
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 math
import base64
import numpy as np

rng = np.random.default_rng()

def _subidx_to_bytes(subidx, n_dims):
    byte_arr = bytearray(math.ceil(n_dims / 8))

    def set_bit(bit_idx):
        byte_idx = bit_idx // 8
        bit_offset = bit_idx % 8
        byte_arr[byte_idx] |= 1 << bit_offset
    for chosen_idx in subidx:
        set_bit(chosen_idx)
    return byte_arr

def subidx_to_tag(subidx, n_dims):
    byte_arr = _subidx_to_bytes(subidx, n_dims)
    return base64.urlsafe_b64encode(byte_arr).decode()

def generate_subidx_tag(n_dims=1003, sample_prop=0.8):
    num_choices = math.floor(n_dims * sample_prop)
    subidx = rng.choice(n_dims, num_choices, replace=False)
    return subidx_to_tag(subidx, n_dims)

ARGS = {
    "n_dims": 100,
    "sample_prop": 0.75,
}
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
get_duplicates_list Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/matching.py
code served (permissive licence) · get_code("6b7a122154d3ef3d")
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

torch.manual_seed(0)

# The function get_duplicates_list takes a list of values and returns duplicates.
# It iterates through values checking if each is in a set, collecting duplicates.
# The function works with any hashable values.

ARGS = {
    "values": [1, 2, 3, 2, 4, 1, 5, 3]
}
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 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_outpath_i Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/visualization/toy_models.py
code served (permissive licence) · get_code("0116669b7cf4cc44")
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 os
import numpy as np

# Analyze the function to understand what arguments it needs:
# - outpath: a string path that contains 'L{p}' pattern
# - p: a number that will be used in string formatting
# - dim: a dimension value
# - snr: a signal-to-noise ratio value

ARGS = {
    "outpath": "/results/model_L5_output.txt",
    "p": 5,
    "dim": 16,
    "snr": 0.1
}
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
make_ldm_images Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/create_ldm_images.py
code served (permissive licence) · get_code("4e1fbc2ec9787f27")
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, ndim 3, 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

# Create small point clouds X and Y
np.random.seed(0)
X = np.random.randn(3, 4)  # 3 points in 4D
Y = np.random.randn(2, 4)  # 2 points in 4D

# Construct pairwise distance matrix for Z (combined X and Y)
Z = np.vstack((X, Y))  # 5 points total
pairwise_Z = np.sqrt(np.sum((Z[:, None, :] - Z[None, :, :]) ** 2, axis=-1))

nb_X = len(X)  # 3

ARGS = {
    'X': X,
    'Y': Y,
    'pairwise_Z': pairwise_Z,
    'nb_X': nb_X,
    'return_thr': False,
    'write_mode': False,
    'filename_X': 'X',
    'filename_Y': 'Y',
    'filename_Z': 'dZ'
}
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 tuple, ndim 3, dtype family float — a tuple 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
plot_cancorrs Ran
ran; our contract draft was wrong, not the code
tyo8/brain_representations/src_py/lindecomp/analyze_CCA_permtests.py
pointer only (licence: NONE) · get_code("23a1ffe2826383ae")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, tyo8/brain_representations 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.

plot_null_cancorrs Ran
ran; our contract draft was wrong, not the code
tyo8/brain_representations/src_py/lindecomp/analyze_CCA_permtests.py
pointer only (licence: NONE) · get_code("1decd4bcdbee36c2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, tyo8/brain_representations 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.

pull_data Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/visualization/bootstrap_dists.py
code served (permissive licence) · get_code("4666812a7fd82270")
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 os
import glob
import pickle
import tempfile
import shutil

# Create a temporary directory structure for testing
temp_dir = tempfile.mkdtemp()

# Create subdirectories with pattern X_*_L2dists
x_dir_1 = os.path.join(temp_dir, 'X_dataset1_L2dists')
x_dir_2 = os.path.join(temp_dir, 'X_dataset2_L2dists')
os.makedirs(x_dir_1, exist_ok=True)
os.makedirs(x_dir_2, exist_ok=True)

# Create pickle files in each directory
sample_data_1 = {'values': [1, 2, 3], 'scores': [0.1, 0.2, 0.3]}
sample_data_2 = {'values': [4, 5, 6], 'scores': [0.4, 0.5, 0.6]}
sample_data_3 = {'values': [7, 8, 9], 'scores': [0.7, 0.8, 0.9]}
sample_data_4 = {'values': [10, 11, 12], 'scores': [1.0, 1.1, 1.2]}

pickle_file_1 = os.path.join(x_dir_1, '*X_*_L2dists.pkl')
pickle_file_2 = os.path.join(x_dir_1, 'Wp_hat_X_vs_Y_result1_L2dists.pkl')
pickle_file_3 = os.path.join(x_dir_2, '*X_*_L2dists.pkl')
pickle_file_4 = os.path.join(x_dir_2, 'Wp_hat_X_vs_Y_result2_L2dists.pkl')

with open(pickle_file_2, 'wb') as f:
    pickle.dump(sample_data_1, f)

with open(pickle_file_4, 'wb') as f:
    pickle.dump(sample_data_3, f)

ARGS = {
    'parent_dir': temp_dir,
    'pattern': '*X_*_L2dists',
    'name_type': 'toy_model',
    'debug': True,
}
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 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
reduce_subidx Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/create_ldm_images.py
code served (permissive licence) · get_code("b8713dc45cfeffae")
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

torch.manual_seed(0)

# Analyze the function:
# - subidx: a list/iterable of indices (integers)
# - N_new: an integer representing a new size/length
# The function filters subidx to keep only indices < N_new,
# then truncates based on a proportion calculation

# subidx should be a list of integers (indices)
subidx = [0, 5, 10, 15, 20, 25, 30]

# N_new should be an integer representing the new maximum index
N_new = 18

ARGS = {
    "subidx": subidx,
    "N_new": N_new,
}
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
run_model Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/visualization/toy_models.py
code served (permissive licence) · get_code("2fb0eddb37453028")
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 (250, 250) 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

# Set random seed for reproducibility
np.random.seed(42)

# Create valid arguments for run_model function
ARGS = {
    'model_type': 'S2uR',
    'N': 250,
    'noise_prop': 0.1,
    'embedding_dim': 3,
    'p': 2,
    'normalize': True,
    'stochastic': False,
    'add_origin': False,
    'plot': False,
    'outpath': None,
    'verbose': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy, scipy — 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 (250, 250) — shape (250, 250) 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
subidx_to_tag Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/generate_subindex.py
code served (permissive licence) · get_code("302d77870ebf0340")
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 math
import base64

def _subidx_to_bytes(subidx, n_dims):
    byte_arr = bytearray(math.ceil(n_dims / 8))

    def set_bit(bit_idx):
        byte_idx = bit_idx // 8
        bit_offset = bit_idx % 8
        byte_arr[byte_idx] |= 1 << bit_offset
    for chosen_idx in subidx:
        set_bit(chosen_idx)
    return byte_arr

def subidx_to_tag(subidx, n_dims):
    byte_arr = _subidx_to_bytes(subidx, n_dims)
    return base64.urlsafe_b64encode(byte_arr).decode()

# Construct valid arguments
ARGS = {
    'subidx': [0, 2, 4, 7],  # list of indices to set as bits
    'n_dims': 16  # number of dimensions
}
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 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
summarize_cycle_registration Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/utils_match/prevalence.py
code served (permissive licence) · get_code("a92f948ccb8664af")
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 numpy as np

# Create a verbose_match_list with dictionaries containing 'affinity' keys
# The function expects a list of match dicts, each with an 'affinity' key
# that contains an array of affinity scores

# Create sample data: 3 features (rows) x 4 bootstrap samples (columns)
verbose_match_list = [
    {'affinity': np.array([0.8, 0.9, 0.7, 0.85])},
    {'affinity': np.array([0.6, 0.75, 0.8, 0.65])},
    {'affinity': np.array([0.9, 0.85, 0.95, 0.88])},
]

ARGS = {
    'verbose_match_list': verbose_match_list
}
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 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
visualize_permtests Ran
ran; our contract draft was wrong, not the code
tyo8/brain_representations/src_py/lindecomp/analyze_CCA_permtests.py
pointer only (licence: NONE) · get_code("8dc57aa55c913eb6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, tyo8/brain_representations 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.

weighted_Wasserstein_dist Ran
ran on a synthesized input
tyo8/interval-matching-precomp_metric/visualization/diagram_distances.py
code served (permissive licence) · get_code("c8122f2adfd18563")
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

# Set up small test data for the weighted_cost_mtx function
# X1 and X2 are persistence diagrams (arrays of shape (n, 2))
# w1 and w2 are weight arrays (arrays of shape (n,))

X1 = np.array([[0.1, 0.5], [0.2, 0.7]])  # 2 points, 2D
X2 = np.array([[0.15, 0.6], [0.25, 0.8], [0.3, 0.9]])  # 3 points, 2D

w1 = np.array([0.8, 0.9])  # weights for X1
w2 = np.array([0.7, 0.85, 0.95])  # weights for X2

ARGS = {
    'X1': X1,
    'X2': X2,
    'w1': w1,
    'w2': w2,
    'p': 2,
    'q': 2,
    'wtfn_type': 'diff',
    'debug': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy, ot — 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
collate_dictlist_data Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/utils_match/collate_tagged_data.py
code served (permissive licence) · get_code("6c82f167af9efd5a")
How it ran

no invocation recorded.

make_pvals_heatmap Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching_bootstrap/visualization/compare_topostats.py
code served (permissive licence) · get_code("376f84588193d7a4")
How it ran

no invocation recorded.

module_distance Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/visualization/diagram_distances.py
code served (permissive licence) · get_code("bb1d9c4a0b57fe04")
How it ran

no invocation recorded.

summarize_diagram Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/utils_match/extract.py
code served (permissive licence) · get_code("ab54762020b6c63d")
How it ran

no invocation recorded.

weighted_cost_mtx Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/visualization/diagram_distances.py
code served (permissive licence) · get_code("72aab4f19441b6ae")
How it ran

no invocation recorded.

write_out Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/visualization/comp_permtest_dists.py
code served (permissive licence) · get_code("711b181f29feeae0")
How it ran

no invocation recorded.

xtr_bars_reps Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/utils_match/extract.py
code served (permissive licence) · get_code("37f29c3f4e096758")
How it ran

no invocation recorded.

xtr_bars_reps_indices Not yet run
unverified (harvested, no recorded run)
tyo8/interval-matching-precomp_metric/utils_match/extract.py
code served (permissive licence) · get_code("53b1da1e64d10b14")
How it ran

no invocation recorded.

Also present in other papers' code

Of 26 samples lifted from this paper's linked repositories, 26 carry a function identity and 0 do not. Of the 26 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 26 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

In neuroimaging, extensive post-processing of resting-state functional MRI (rfMRI) data is necessary for its application and investigation in relation to brain-behavior associations. Such post-processing is used to derive brain representations, lower dimensional feature sets used for brain-behavior association studies. A brain representation involves a choice of dimension reduction (a parcellation into regions or networks) and a choice of feature type, such as spatial topography, connectivity matrix, amplitude. However, widespread variability in rfMRI brain representations has hindered both reproducibility and knowledge accumulation across the field. Brain representation choice effects measurements of inter-subject variability, which muddies the comparison and integration of findings. We leveraged persistent homology on the subject-space topologies induced by 34 different brain representations to enable direct comparison of brain representations in the context of individual differences. Our findings reveal the importance of considering feature type when comparing results derived from different brain representations, suggesting best practices for assessing the replicability and generalizability of brain-behavior research in rfMRI data.

For agents

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

get_harvested_code_for_paper("2306.13802")
get_code_for_paper("2306.13802")
have("2306.13802")

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

curl https://syntology.ai/api/ran/2306.13802.json

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

[![Syntology run record](https://syntology.ai/api/ran/2306.13802.svg)](https://syntology.ai/paper/2306.13802)

Connect an agent — have() is free.