SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1606.07892 · 2016

Large-Scale Kernel Methods for Independence Testing

arXiv · PDF · Open in the Atlas

Code that ran

13 of 17 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 13 have no contract verdict; 0 failures of our instrument; 4 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/1606.07892.json · badge: /api/ran/1606.07892.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
IBM/SIC community (classifier label)
classifier basis not recorded on the link
13 ran · 4 no recorded run
FunctionStatusWhere it lives
compute_fdr Ran
ran on a synthesized input
IBM/SIC/stattests.py
code served (permissive licence) · get_code("d81b7960091f43e3")
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 random seed for reproducibility
np.random.seed(0)

# Define arguments for compute_fdr
eta_sortix = np.array([0, 2, 5, 8, 1, 3])  # indices of top features sorted by some criterion
ground_truth = np.array([0, 2, 5])  # indices of true features
total_features = 20  # total number of features
axis = None  # axis parameter for sum operations
cut_off = 40  # cutoff for number of features to consider

ARGS = {
    "eta_sortix": eta_sortix,
    "ground_truth": ground_truth,
    "total_features": total_features,
    "axis": axis,
    "cut_off": cut_off,
}
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 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
compute_mse Ran
ran on a synthesized input
IBM/SIC/SIC_imports.py
code served (permissive licence) · get_code("6a9bfdd55c282718")
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 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 torch.nn as nn
import torch.nn.functional as F

torch.manual_seed(0)

# Create a simple neural network that takes input and outputs predictions
class SimpleNet(nn.Module):
    def __init__(self, input_dim, output_dim):
        super().__init__()
        self.linear = nn.Linear(input_dim, output_dim)
    
    def forward(self, x):
        return self.linear(x)

# Create network
net = SimpleNet(input_dim=4, output_dim=1)

# Create a simple dataloader as a list of (data, targets) tuples
# Batch size 2, input dimension 4, output dimension 1
dataloader = [
    (torch.randn(2, 4), torch.randn(2, 1)),
    (torch.randn(2, 4), torch.randn(2, 1)),
]

# Normalization parameters for targets
targets_mu = 0.0
targets_sd = 1.0

ARGS = {
    'dataloader': dataloader,
    'net': net,
    'targets_mu': targets_mu,
    'targets_sd': targets_sd,
}
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 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
compute_objective_terms Ran
ran on a synthesized input
IBM/SIC/SIC_imports.py
code served (permissive licence) · get_code("d07e6ab7af15ef76")
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 torch.nn as nn

torch.manual_seed(0)

class SimpleNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x, y):
        # x shape: (batch, 4), y shape: (batch, 5)
        # Return shape: (batch,)
        return self.linear(x).sum(dim=1)

net = SimpleNet()

# Create data: (x, y) tuple
# x: batch_size=2, feature_dim=4
# y: batch_size=2, label_dim=5
x = torch.randn(2, 4)
y = torch.randn(2, 5)
data = (x, y)

ARGS = {
    "data": data,
    "net": net,
    "need_penalty_terms": True
}
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
describe Ran
ran on a synthesized input
IBM/SIC/utils.py
code served (permissive licence) · get_code("d204ab0c6f9c28e9")
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 torch
import numpy as np

torch.manual_seed(0)

# Test with a numpy array
numpy_array = np.random.randn(2, 4).astype(np.float32)

# Test with a torch tensor
torch_tensor = torch.randn(3, 5)

# Create ARGS dict with both types to demonstrate the function works
ARGS = {
    't': numpy_array
}
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 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
dlist Ran
ran on a synthesized input
IBM/SIC/plot_results.py
code served (permissive licence) · get_code("014bc653675795b0")
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 (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

# The function dlist takes:
# 1. key: a string to look up in dictionaries
# 2. dat: a list of dictionaries

# Create a list of dictionaries with varying content
dat = [
    {'name': 'alice', 'age': 30, 'score': 95},
    {'name': 'bob', 'score': 87},  # missing 'age'
    {'name': 'charlie', 'age': 25, 'score': 92},
    {'score': 88},  # missing 'name' and 'age'
]

ARGS = {
    'key': 'score',
    'dat': dat
}
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 (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
get_data Ran
ran on a synthesized input
IBM/SIC/plot_results.py
code served (permissive licence) · get_code("ea4d799e3e40806a")
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

# Construct minimal valid arguments for get_data

# select_dict: list of dicts, each with one key-value pair used for filtering
select_dict = [
    {'model_type': 'lstm'},
    {'model_type': 'transformer'}
]

# key_list: list of keys to extract from filtered data
key_list = ['loss', 'accuracy']

# DAT: list of lists of dictionaries (data to filter and extract from)
DAT = [
    [
        {'loss': 0.5, 'accuracy': 0.8},
        {'loss': 0.3, 'accuracy': 0.9}
    ],
    [
        {'loss': 0.4, 'accuracy': 0.85},
        {'loss': 0.2, 'accuracy': 0.95}
    ],
    [
        {'loss': 0.6, 'accuracy': 0.75},
        {'loss': 0.4, 'accuracy': 0.88}
    ]
]

# ARGS: list of dictionaries with filtering keys
ARGS = [
    {'model_type': 'lstm', 'lr': 0.001},
    {'model_type': 'transformer', 'lr': 0.0001},
    {'model_type': 'cnn', 'lr': 0.01}
]

ARGS = {
    'select_dict': select_dict,
    'ARGS': ARGS,
    'key_list': key_list,
    'DAT': DAT
}
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
get_devices Ran
ran on a synthesized input
IBM/SIC/utils.py
code served (permissive licence) · get_code("290087b87df2ac2e")
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 1, dtype family float; no output shape is stated. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 = {
    'cuda_device': 'cpu',
    'seed': 1,
}
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, ndim 1, 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
init_optimizerD Ran
ran on a synthesized input
IBM/SIC/modules/models.py
code served (permissive licence) · get_code("74d8b31b37ed27e1")
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 Adam. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create a simple discriminator model
class SimpleDiscriminator(nn.Module):
    def __init__(self):
        super().__init__()
        self.features = nn.Sequential(
            nn.Linear(4, 8),
            nn.ReLU(),
            nn.Linear(8, 16)
        )
        self.last_linear = nn.Linear(16, 1)
    
    def forward(self, x):
        x = self.features(x)
        x = self.last_linear(x)
        return x

# Create a simple config object with the required attributes
class OptConfig:
    def __init__(self):
        self.DiscArch = 'concat_first'
        self.lrD = 0.0002
        self.beta1 = 0.5
        self.beta2 = 0.999
        self.wdecay = 0.0

# Create instances
opt = OptConfig()
D = SimpleDiscriminator()

ARGS = {
    'opt': opt,
    'D': D,
    'train_last_layer_only': False
}
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 Adam — a Adam output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Adam 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
label_axis Ran
ran on a synthesized input
IBM/SIC/plot_results.py
code served (permissive licence) · get_code("cedc42f38350b7b6")
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 Axes. 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 matplotlib.pyplot as plt
import numpy as np

# Create a minimal matplotlib axes object
fig, ax = plt.subplots()

# Plot some dummy data to initialize the axes properly
ax.plot([0, 1, 2], [0, 0.5, 1])

# Define labels for the x-axis
labels = ['Label A', 'Label B', 'Label C']

# Define x-axis positions (where labels should be placed)
xpos = [0.2, 0.5, 0.8]

# Define y-axis position for the spine offset
ypos = 10

# Define fontsize
fontsize = 14

# Define target FDR threshold
target_fdr = 0.05

ARGS = {
    'ax': ax,
    'labels': labels,
    'xpos': xpos,
    'ypos': ypos,
    'fontsize': fontsize,
    'target_fdr': target_fdr
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
matplotlib, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Axes — a Axes output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Axes 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
linreg_reconstruct Ran
ran on a synthesized input
IBM/SIC/modules/linregression.py
code served (permissive licence) · get_code("8e090872f4c05874")
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 LinRegression. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create a small input tensor of size (batch_size=2, num_features=4)
X = torch.randn(2, 4)

# idx_feature is an integer index selecting which feature to reconstruct
idx_feature = 1

# eps is a float regularization parameter
eps = 1e-06

ARGS = {
    "X": X,
    "idx_feature": idx_feature,
    "eps": eps
}
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 LinRegression — a LinRegression output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a LinRegression 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
minibatch Ran
ran on a synthesized input
IBM/SIC/SIC_imports.py
code served (permissive licence) · get_code("2f5a64db1f7ac832")
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 tensors for x and y
x = torch.randn(10, 4, requires_grad=False)  # 10 samples, 4 features
y = torch.randn(10, 3, requires_grad=False)  # 10 samples, 3 targets

# data is a tuple of (x, y)
data = (x, y)

# batch_size can be an integer or None/0
batch_size = 2

# requires_grad is a boolean
requires_grad = True

ARGS = {
    "data": data,
    "batch_size": batch_size,
    "requires_grad": requires_grad,
}
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
test_generator Ran
ran on a synthesized input
IBM/SIC/modules/generators.py
code served (permissive licence) · get_code("a919d57070b61232")
How it ran

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

import torch
import torch.nn as nn

torch.manual_seed(0)

# Create a minimal generator model that satisfies the function's requirements
class MinimalGenerator(nn.Module):
    def __init__(self, feature_dim=4):
        super().__init__()
        self.feature_dim = feature_dim
        self.linear = nn.Linear(feature_dim + 1, feature_dim)
    
    def get_targets(self, data, labels):
        # data shape: (batch_size, feature_dim)
        # labels shape: (batch_size,)
        # Return targets with same shape as outputs
        return torch.randn(data.shape[0], self.feature_dim)
    
    def sample_features(self, data_and_labels):
        # data_and_labels is a tuple of (data, labels)
        data, labels = data_and_labels
        # data shape: (batch_size, feature_dim)
        # labels shape: (batch_size,)
        # Concatenate labels as float and pass through linear layer
        labels_float = labels.float().unsqueeze(1)
        combined = torch.cat([data, labels_float], dim=1)
        return self.linear(combined)

# Create a minimal criterion (loss function)
criterion = nn.MSELoss()

# Create a minimal dataloader
# Each element should be a tuple where the first element is a tensor
# The function does data[0].to(device), so we need (tensor, dummy)
batch_size = 2
feature_dim = 4
dataloader = [
    (torch.randn(batch_size, feature_dim), None),
    (torch.randn(batch_size, feature_dim), None),
]

# Create the generator
generator = MinimalGenerator(feature_dim=feature_dim)

ARGS = {
    "generator": generator,
    "criterion": criterion,
    "dataloader": dataloader,
    "test_all_features": False,
}
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 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
train_generator Ran
ran on a synthesized input
IBM/SIC/modules/generators.py
code served (permissive licence) · get_code("adbde4eae22a3f8f")
How it ran

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

import torch
import torch.nn as nn
import torch.optim as optim

torch.manual_seed(0)

# Define a minimal Generator class that satisfies the function's usage
class MinimalGenerator(nn.Module):
    def __init__(self, num_features=4):
        super().__init__()
        self.num_features = num_features
        self.linear = nn.Linear(num_features, num_features)
    
    def training_loss(self, data, labels):
        # Minimal loss computation using data and labels
        output = self.linear(data)
        loss = output.sum() + labels.float().sum()
        return loss

# Create generator instance
generator = MinimalGenerator(num_features=4)

# Create a minimal dataloader (list of batches)
# Each batch is a tensor of shape (batch_size, num_features)
dataloader = [
    torch.randn(2, 4),
    torch.randn(2, 4),
]
# Wrap to have __len__ and __iter__
class SimpleDataLoader:
    def __init__(self, data):
        self.data = data
        self.dataset = self
        self.length = len(data)
    
    def __len__(self):
        return self.length
    
    def __iter__(self):
        return iter(self.data)

dataloader = SimpleDataLoader(dataloader)

# Create optimizer
optimizer = optim.SGD(generator.parameters(), lr=0.01)

# Features list (optional, will be set to range if None)
features_list = [0, 1, 2, 3]

# Log times
log_times = 2

ARGS = {
    "generator": generator,
    "dataloader": dataloader,
    "optimizer": optimizer,
    "features_list": features_list,
    "log_times": log_times,
}
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 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
init_D Not yet run
unverified (harvested, no recorded run)
IBM/SIC/modules/models.py
code served (permissive licence) · get_code("fb54dc11aed4cbf8")
How it ran

no invocation recorded.

make_data_parallel Not yet run
unverified (harvested, no recorded run)
IBM/SIC/utils.py
code served (permissive licence) · get_code("0b8190dcac4fd585")
How it ran

no invocation recorded.

risk_model_fn Not yet run
unverified (harvested, no recorded run)
IBM/SIC/run_baselines.py
code served (permissive licence) · get_code("6c42a9c801164386")
How it ran

no invocation recorded.

train_sklearn_model Not yet run
unverified (harvested, no recorded run)
IBM/SIC/run_baselines.py
code served (permissive licence) · get_code("8638295ef8c7c717")
How it ran

no invocation recorded.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 17 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 2 later) and 15 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
get_devices 1 other paper
1 later
2009.14148 (2020-09)
Unbalanced Sobolev Descent
make_data_parallel 1 other paper
1 later
2009.14148 (2020-09)
Unbalanced Sobolev Descent

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

Representations of probability measures in reproducing kernel Hilbert spaces provide a flexible framework for fully nonparametric hypothesis tests of independence, which can capture any type of departure from independence, including nonlinear associations and multivariate interactions. However, these approaches come with an at least quadratic computational cost in the number of observations, which can be prohibitive in many applications. Arguably, it is exactly in such large-scale datasets that capturing any type of dependence is of interest, so striking a favourable tradeoff between computational efficiency and test performance for kernel independence tests would have a direct impact on their applicability in practice. In this contribution, we provide an extensive study of the use of large-scale kernel approximations in the context of independence testing, contrasting block-based, Nystrom and random Fourier feature approaches. Through a variety of synthetic data experiments, it is demonstrated that our novel large scale methods give comparable performance with existing methods whilst using significantly less computation time and memory.

For agents

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

get_harvested_code_for_paper("1606.07892")
get_code_for_paper("1606.07892")
have("1606.07892")

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

curl https://syntology.ai/api/ran/1606.07892.json

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

[![Syntology run record](https://syntology.ai/api/ran/1606.07892.svg)](https://syntology.ai/paper/1606.07892)

Connect an agent — have() is free.