SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.03806 · 2020

Leveraging the Feature Distribution in Transfer-based Few-Shot Learning

arXiv · PDF · Open in the Atlas

Code that ran

22 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 12 have no contract verdict; 9 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/2006.03806.json · badge: /api/ran/2006.03806.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
yhu01/PT-MAP canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
mbonto/fewshot_neuroimaging_classification community (classifier label)
classifier basis not recorded on the link
14 ran · 1 no recorded run
yhu01/bms reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
xiangyu8/PT-MAP-sf reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
allenhaozhu/ease extension (classifier label)
classifier basis: classifier_2025
0 ran · 3 no recorded run
FunctionStatusWhere it lives
QRreduction Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yhu01/PT-MAP/test_standard.py
pointer only (licence: NONE) · get_code("c2a29a7f0baf902f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, yhu01/PT-MAP 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.

augment_image Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/data.py
code served (permissive licence) · get_code("18f856fdae4a4061")
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 3, dtype family float, shape (2, 4, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create a small numpy image array (2, 4, 3) - batch_size=2, height=4, width=3
image = np.random.rand(2, 4, 3).astype(np.float32)

# k parameter - integer representing number of augmentations or similar
k = 2

# channels - number of image channels
channels = 3

# augment_bool - boolean flag to enable/disable augmentation
augment_bool = True

# args - configuration dictionary or object with augmentation parameters
args = {
    'augment_type': 'random_flip',
    'intensity': 0.5,
    'enable_rotation': True,
}

# dataset_name - string identifier for the dataset
dataset_name = 'cifar10'

ARGS = {
    'image': image,
    'k': k,
    'channels': channels,
    'augment_bool': augment_bool,
    'args': args,
    'dataset_name': dataset_name,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 3, dtype family float, shape (2, 4, 3) — shape (2, 4, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
batch_shape Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/SimpleShot/functions.py
code served (permissive licence) · get_code("c160c8210094d8c7")
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 3, dtype family float, shape (2, 1, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function takes model_name (str) and x (tensor)
# It performs squeeze/unsqueeze operations based on model_name
# The tensor needs to have at least 2 dimensions to be meaningful for these operations

# Create a small tensor with shape (batch=2, channels=1, features=4)
x = torch.ones(2, 1, 4)

ARGS = {
    'model_name': 'Conv1d',
    'x': x
}
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 3, dtype family float, shape (2, 1, 4) — shape (2, 1, 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
centerDatas Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yhu01/PT-MAP/test_standard.py
pointer only (licence: NONE) · get_code("c317909266ed4d67")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, yhu01/PT-MAP 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.

centerDatas Ran
ran, honoured the contract we drafted
yhu01/bms/test_standard_bms.py
pointer only (licence: NONE) · get_code("600b828ba2526da0")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, yhu01/bms 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.

centerDatas Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/SimpleShot/PT_plus_MAP.py
code served (permissive licence) · get_code("1586cba5790084af")
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 3, dtype family float, shape (2, 5, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function expects:
# - datas: a 3D tensor with shape (batch, n_samples, features)
#   It indexes datas[:, :n_lsamples] and datas[:, n_lsamples:]
#   so n_samples must be > n_lsamples
# - n_lsamples: an integer indicating split point in second dimension

# Create a small 3D tensor: (batch=2, n_samples=5, features=4)
# This allows n_lsamples to be any value from 1-4
datas = torch.randn(2, 5, 4)

# Set n_lsamples to split the samples dimension
n_lsamples = 3

ARGS = {
    "datas": datas,
    "n_lsamples": n_lsamples,
}
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 3, dtype family float, shape (2, 5, 4) — shape (2, 5, 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
compute_accuracy Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/SimpleShot/functions.py
code served (permissive licence) · get_code("a280c947612b8a09")
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 int, ndim 0, dtype family int, 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

torch.manual_seed(0)

# Create outputs tensor: batch_size=2, num_classes=4
outputs = torch.randn(2, 4)

# Create target labels tensor: batch_size=2
y = torch.tensor([0, 2])

ARGS = {
    "outputs": outputs,
    "y": y,
}
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 int, ndim 0, dtype family int, 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_normalized_adjacency Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/meta_neural_network_architectures.py
code served (permissive licence) · get_code("aa05f8300c098828")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (4, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

torch.manual_seed(0)

# Create a small adjacency matrix (4x4)
# Values should mostly be in range [0, 1] with some > 0.945 to test thresholding
adj_matrix = np.array([
    [0.5, 0.95, 0.3, 0.2],
    [0.95, 0.6, 0.8, 0.1],
    [0.3, 0.8, 0.7, 0.96],
    [0.2, 0.1, 0.96, 0.4]
], dtype=np.float32)

ARGS = {
    "adj_matrix": adj_matrix,
    "threshold": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (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
extract_top_level_dict Ran
ran; our contract draft was wrong, not the code
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/meta_neural_network_architectures.py
pointer only (licence: MIT); the stored copy get_code serves, from baiksung/MeTAL, is pointer only (licence: NONE) · get_code("e6cee89c00846aec")
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 baiksung/MeTAL, 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.

fill_template Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/script_generation_tools/generate_scripts.py
code served (permissive licence) · get_code("d25adfe464700ae6")
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
from copy import copy

# Construct valid arguments for fill_template function

# The function takes a template_list (list of strings), execution_script (string),
# and experiment_config (string). It performs string replacements on the last element
# of template_list and joins all elements together.

template_list = [
    "#!/usr/bin/env python\n",
    "import sys\n",
    "# Execution script: $execution_script$\n",
    "# Config: $experiment_config$\n",
    "print('Running with config')\n"
]

execution_script = "train_model.py"

experiment_config = "config_v1.yaml"

ARGS = {
    'template_list': template_list,
    'execution_script': execution_script,
    'experiment_config': experiment_config
}
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 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_combinations Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/script_generation_tools/generate_configs.py
code served (permissive licence) · get_code("2f7232018db3fa2e")
How it ran

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

import torch
import numpy as np
from collections import namedtuple

torch.manual_seed(0)

# The function expects a config namedtuple where each field is iterable
# The function iterates over config._asdict().items() and treats each value as a list of choices
# It then generates all combinations

config = namedtuple('config', 'dataset_name num_classes samples_per_class target_samples_per_class batch_size train_update_steps val_update_steps init_inner_loop_learning_rate load_into_memory learnable_bn_beta learnable_bn_gamma conv_padding num_filters experiment_name')

# Create a config object where each field is a list of possible values
test_config = config(
    dataset_name=['cifar10', 'mnist'],
    num_classes=[5, 10],
    samples_per_class=[4, 8],
    target_samples_per_class=[2, 4],
    batch_size=[16, 32],
    train_update_steps=[5, 10],
    val_update_steps=[3, 5],
    init_inner_loop_learning_rate=[0.001, 0.01],
    load_into_memory=[True, False],
    learnable_bn_beta=[True, False],
    learnable_bn_gamma=[True, False],
    conv_padding=[1, 2],
    num_filters=[32, 64],
    experiment_name=['exp1', 'exp2']
)

ARGS = {
    'config': test_config
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_transforms_for_dataset Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/data.py
code served (permissive licence) · get_code("40f79117fc32f74c")
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 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Based on the function, it takes:
# - dataset_name: str (checked for 'IBC' substring)
# - args: appears to be unused in this function
# - k: appears to be unused in this function

# For a dataset_name that triggers the 'IBC' condition
dataset_name = "IBC_dataset"

# args can be any object since it's not used in the function body
args = {}

# k can be any value since it's not used in the function body
k = 1

ARGS = {
    "dataset_name": dataset_name,
    "args": args,
    "k": k
}
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 2, 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
load_pickle Ran
ran; our contract draft was wrong, not the code
yhu01/bms/test_standard_bms.py
pointer only (licence: NONE) · get_code("3fffadb9535ea4c6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, yhu01/bms 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.

load_template Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/script_generation_tools/generate_configs.py
code served (permissive licence) · get_code("27fabd65989ead1f")
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 tempfile
import os

# Create a temporary file with some template content
temp_dir = tempfile.mkdtemp()
temp_filepath = os.path.join(temp_dir, "template.txt")

template_content = """
This is a sample template file.
It contains some text that will be read by load_template.
Line 3: More content here.
"""

with open(temp_filepath, 'w') as f:
    f.write(template_content)

ARGS = {
    'filepath': temp_filepath
}
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
load_template Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/script_generation_tools/generate_scripts.py
code served (permissive licence) · get_code("4a41646d1d2da5a6")
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 tempfile
import os

# Create a temporary file with sample template content
temp_dir = tempfile.mkdtemp()
temp_filepath = os.path.join(temp_dir, "template.txt")

with open(temp_filepath, mode='w') as f:
    f.write("Line 1 of template\n")
    f.write("Line 2 of template\n")
    f.write("Line 3 of template\n")

ARGS = {
    "filepath": temp_filepath
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
propagation Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/meta_neural_network_architectures.py
code served (permissive licence) · get_code("0e4987f8a7cbc4cb")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# features: tensor of size (batch, num_features)
# Using batch=2, num_features=4
features = torch.randn(2, 4)

# graph: adjacency matrix of size (num_features, num_features)
# Using num_features=4, so 4x4 matrix
graph = torch.randn(4, 4)

# k: number of propagation steps
k = 2

ARGS = {
    "features": features,
    "graph": graph,
    "k": k
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
reloadRuns Ran
ran; our contract draft was wrong, not the code
yhu01/bms/test_standard_bms.py
pointer only (licence: NONE) · get_code("ca2929dfbe4faf3e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, yhu01/bms 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.

scaleEachUnitaryDatas Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yhu01/PT-MAP/test_standard.py
pointer only (licence: NONE) · get_code("7bf4f43bdbd661d4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, yhu01/PT-MAP 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.

set_torch_seed Ran
ran; our contract draft was wrong, not the code
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/few_shot_learning_system.py
code served (permissive licence) · get_code("5c44538154dad8a7")
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 RandomState. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np
import torch

ARGS = {
    "seed": 42
}
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 RandomState — a RandomState output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a RandomState 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
train Ran
ran on a synthesized input
mbonto/fewshot_neuroimaging_classification/SimpleShot/functions.py
code served (permissive licence) · get_code("d5f7aba697839f38")
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
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self, model_name):
        super().__init__()
        self.name = model_name
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create a small model
model = SimpleModel('MLP')

# Create criterion and optimizer
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)

# Create a small data loader (list of (x, y) tuples)
# For MLP: x should be (batch, features)
# batch_shape will squeeze dim 1, so we pass (batch, 1, features)
data_loader = [
    (torch.ones(2, 1, 4), torch.tensor([0, 1])),
    (torch.ones(2, 1, 4), torch.tensor([1, 0])),
]

# use_cuda flag
use_cuda = False

ARGS = {
    'model': model,
    'criterion': criterion,
    'optimizer': optimizer,
    'data_loader': data_loader,
    'use_cuda': use_cuda
}
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
train_manifold_mixup Ran
ran; our contract draft was wrong, not the code
xiangyu8/PT-MAP-sf/train_dct.py
pointer only (licence: NONE) · get_code("16aa65e2f94bd50a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, xiangyu8/PT-MAP-sf 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.

train_rotation Ran
ran; our contract draft was wrong, not the code
xiangyu8/PT-MAP-sf/train_dct.py
pointer only (licence: NONE) · get_code("085fb3f0b676ad6a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, xiangyu8/PT-MAP-sf 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.

ClassesInRun Not yet run
unverified (harvested, no recorded run)
allenhaozhu/ease/FSLTask.py
pointer only (licence: NONE) · get_code("9ae865b05537face")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, allenhaozhu/ease 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. No invocation recorded.

GenerateRun Not yet run
unverified (harvested, no recorded run)
allenhaozhu/ease/FSLTask.py
pointer only (licence: NONE) · get_code("a60c716cc71e54f7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, allenhaozhu/ease 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. No invocation recorded.

GenerateRunSet Not yet run
unverified (harvested, no recorded run)
allenhaozhu/ease/FSLTask.py
pointer only (licence: NONE) · get_code("56ad4643b6fbdddd")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, allenhaozhu/ease 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. No invocation recorded.

fill_template Not yet run
unverified (harvested, no recorded run)
mbonto/fewshot_neuroimaging_classification/MAML_plus_plus/script_generation_tools/generate_configs.py
code served (permissive licence) · get_code("6ef99887903cd30a")
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, 10 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 7 later) and 16 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

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

FunctionAlso present inEarliest-datedOthers named
set_torch_seed 6 other papers
5 dated earlier, 1 later · 1 only as an assembly brick
1703.03400 (2017-03)
Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
1810.09502 (2018-10)
How to train your MAML
1905.10295 (2019-05)
Learning to learn via Self-Critique
and 3 more not named here
ClassesInRun 2 other papers
2 later
Lazarou_Iterative_Label_Cleaning_for_Transductive_and_Semi-Supervised_Few-Shot_Learning_ICCV_2021_paper (2021 (year only))
Iterative label cleaning for transductive and semi-supervised few-shot learning
2202.02471 (2022-02)
CLUSTER-INDUCED VORONOI DIAGRAMS: A GEOMETRIC APPROACH
GenerateRun 1 other paper
1 later
Lazarou_Iterative_Label_Cleaning_for_Transductive_and_Semi-Supervised_Few-Shot_Learning_ICCV_2021_paper (2021 (year only))
Iterative label cleaning for transductive and semi-supervised few-shot learning
GenerateRunSet 1 other paper
1 later
Lazarou_Iterative_Label_Cleaning_for_Transductive_and_Semi-Supervised_Few-Shot_Learning_ICCV_2021_paper (2021 (year only))
Iterative label cleaning for transductive and semi-supervised few-shot learning
QRreduction 1 other paper
1 later
Lazarou_Iterative_Label_Cleaning_for_Transductive_and_Semi-Supervised_Few-Shot_Learning_ICCV_2021_paper (2021 (year only))
Iterative label cleaning for transductive and semi-supervised few-shot learning
extract_top_level_dict 1 other paper
1 later
2110.03909 (2021-10)
Meta-Learning with Task-Adaptive Loss Function for Few-Shot Learning
fill_template 1 other paper
1 dated earlier
1905.10295 (2019-05)
Learning to learn via Self-Critique
load_template 1 other paper
1 dated earlier
1905.10295 (2019-05)
Learning to learn via Self-Critique
load_template 1 other paper
1 dated earlier
1905.10295 (2019-05)
Learning to learn via Self-Critique
scaleEachUnitaryDatas 1 other paper
1 later
Lazarou_Iterative_Label_Cleaning_for_Transductive_and_Semi-Supervised_Few-Shot_Learning_ICCV_2021_paper (2021 (year only))
Iterative label cleaning for transductive and semi-supervised few-shot learning

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

Few-shot classification is a challenging problem due to the uncertainty caused by using few labelled samples. In the past few years, many methods have been proposed to solve few-shot classification, among which transfer-based methods have proved to achieve the best performance. Following this vein, in this paper we propose a novel transfer-based method that builds on two successive steps: 1) preprocessing the feature vectors so that they become closer to Gaussian-like distributions, and 2) leveraging this preprocessing using an optimal-transport inspired algorithm (in the case of transductive settings). Using standardized vision benchmarks, we prove the ability of the proposed methodology to achieve state-of-the-art accuracy with various datasets, backbone architectures and few-shot settings.

For agents

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

get_harvested_code_for_paper("2006.03806")
get_code_for_paper("2006.03806")
have("2006.03806")

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

curl https://syntology.ai/api/ran/2006.03806.json

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

[![Syntology run record](https://syntology.ai/api/ran/2006.03806.svg)](https://syntology.ai/paper/2006.03806)

Connect an agent — have() is free.