SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1906.06148 · 2019

A Partially Reversible U-Net for Memory-Efficient Volumetric Image Segmentation

arXiv · PDF · Open in the Atlas

Code that ran

15 of 15 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 15 have no contract verdict; 0 failures of our instrument; 0 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/1906.06148.json · badge: /api/ran/1906.06148.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
RobinBruegger/PartiallyReversibleUnet canonical (classifier label)
classifier basis: census
15 ran · 0 no recorded run
FunctionStatusWhere it lives
convert_to_onehot Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/utils.py
code served (permissive licence) · get_code("655b6c41d66f6e08")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (2, 3, 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 numpy as np

# Create a label map with shape (2, 3) containing label indices
lblmap = np.array([[0, 1, 2],
                   [1, 0, 2]], dtype=np.int32)

# Number of labels
nlabels = 3

ARGS = {
    "lblmap": lblmap,
    "nlabels": nlabels
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (2, 3, 3) — shape (2, 3, 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
cropToSize Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/augmentation.py
code served (permissive licence) · get_code("490170484880eb83")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (6, 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 test image with shape (height, width, channels)
# Using distinct dimensions: 10x8x3 to make shapes unambiguous
image = np.random.rand(10, 8, 3).astype(np.float32)

# Define target size as (height, width)
targetSize = (6, 4)

ARGS = {
    "image": image,
    "targetSize": targetSize,
}
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 ndarray, ndim 3, dtype family float, shape (6, 4, 3) — shape (6, 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
crop_volume_allDim Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/brats18_data_loader.py
code served (permissive licence) · get_code("6bdbaa82995e50da")
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 4D image array (height, width, depth, channels)
# Shape: (5, 6, 7, 1) - small dimensions to keep it tiny
# The function expects image > 0 to find non-zero coordinates
image = np.zeros((5, 6, 7, 1), dtype=np.float32)
# Add some non-zero values in the middle to create a valid crop region
image[1:4, 2:5, 3:6, 0] = 1.0

# Create a mask with the same spatial dimensions
mask = np.ones((5, 6, 7), dtype=np.float32)

ARGS = {
    "image": image,
    "mask": mask,
}
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
crop_volume_allDim Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/brats18_validation_data_loader.py
code served (permissive licence) · get_code("a4685fdb16dd025a")
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 small 4D image volume with shape (height, width, depth, channels)
# The image has some non-zero values in the center and zeros on the edges
image = np.zeros((8, 6, 5, 3), dtype=np.float32)
# Add some non-zero values in a smaller region
image[2:6, 1:5, 1:4, :] = np.random.rand(4, 4, 3, 3)

# Create a corresponding mask with the same spatial dimensions (no channel dimension)
mask = np.zeros((8, 6, 5), dtype=np.float32)
mask[2:6, 1:5, 1:4] = 1.0

ARGS = {
    "image": image,
    "mask": mask,
}
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
dice Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/bratsUtils.py
code served (permissive licence) · get_code("bc6738cdc2e9cdac")
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)

# The dice function expects:
# - pred: a tensor that will be compared against 0.5 and converted to float
# - target: a tensor to compare against

# Looking at the function body:
# - pred is used with (pred > 0.5) to create a binary tensor
# - Both pred and target are passed to softDice
# - softDice performs operations: sum(dim=(1,2,3)), sum(), so need 4D tensors
# - The tensors need batch dimension and spatial dimensions

# Create small 4D tensors (batch=2, channels=1, height=4, width=3)
batch_size = 2
channels = 1
height = 4
width = 3

pred = torch.rand(batch_size, channels, height, width)
target = torch.randint(0, 2, (batch_size, channels, height, width)).float()

ARGS = {
    "pred": pred,
    "target": target
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type 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
diceLoss Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/bratsUtils.py
code served (permissive licence) · get_code("ad0498c6f103c936")
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

torch.manual_seed(0)

# Create small tensors for pred and target
# pred: predictions, should be float values (logits or probabilities)
# target: ground truth binary labels
# Shape: (batch=2, channels=1, height=4, width=3)
pred = torch.rand(2, 1, 4, 3, dtype=torch.float32)
target = torch.randint(0, 2, (2, 1, 4, 3), dtype=torch.float32)

ARGS = {
    'pred': pred,
    'target': target,
    'nonSquared': 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 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
getChannelsAtIndex Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/experiments/noNewReversible.py
code served (permissive licence) · get_code("cfad01770826b0d7")
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
import numpy as np

CHANNELS = [60, 120, 240, 360, 480]

def getChannelsAtIndex(index):
    if index < 0:
        index = 0
    if index >= len(CHANNELS):
        index = len(CHANNELS) - 1
    return CHANNELS[index]

# Test with a valid index
ARGS = {
    "index": 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type 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
ncc Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/utils.py
code served (permissive licence) · get_code("64238175a9a6b6ce")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (1,) 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

# Create simple numpy arrays for testing
a = np.array([1.0, 2.0, 3.0, 4.0, 5.0])
v = np.array([2.0, 3.0, 4.0, 5.0, 6.0])

ARGS = {
    "a": a,
    "v": v,
    "zero_norm": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (1,) — shape (1,) 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
norm_l2 Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/utils.py
code served (permissive licence) · get_code("40982e09ede903ad")
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

# The function norm_l2 takes two parameters: a and v
# Both are used with .flatten(), so they should be array-like objects
# The function performs mathematical operations on them (mean, std, etc.)
# so they should be numpy arrays or convertible to them

# Create two distinct numpy arrays for testing
a = np.array([[1.0, 2.0, 3.0, 4.0],
              [5.0, 6.0, 7.0, 8.0]])

v = np.array([[0.5, 1.5, 2.5, 3.5],
              [4.5, 5.5, 6.5, 7.5]])

ARGS = {
    "a": a,
    "v": v
}
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
normalise_image Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/brats18_data_loader.py
code served (permissive licence) · get_code("9146f30cacc79517")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (4, 3, 2) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create a small test image with some zero pixels and some nonzero pixels
# Shape: (height, width, channels) = (4, 3, 2)
image = np.array([
    [[1.0, 2.0], [3.0, 4.0], [5.0, 6.0]],
    [[0.0, 0.0], [7.0, 8.0], [9.0, 10.0]],
    [[11.0, 12.0], [0.0, 0.0], [13.0, 14.0]],
    [[15.0, 16.0], [17.0, 18.0], [0.0, 0.0]]
], dtype=np.float32)

ARGS = {
    "image": image
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (4, 3, 2) — shape (4, 3, 2) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
normalise_image Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/brats18_validation_data_loader.py
code served (permissive licence) · get_code("796feb0e31cfb904")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (4, 3, 2) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Create a small test image with shape (height, width, channels)
# Using a 4x3x2 image: height=4, width=3, channels=2
image = np.random.rand(4, 3, 2).astype(np.float32) * 100

# Add some zeros to test the nonzero pixel standardization
image[0, 0, :] = 0
image[1, 1, :] = 0

ARGS = {
    "image": image
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (4, 3, 2) — shape (4, 3, 2) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
padToSize Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/augmentation.py
code served (permissive licence) · get_code("abd3eebccec9f78b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family float, shape (8, 10, 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 numpy as np

# Create a small 3D image array (height, width, channels)
image = np.random.rand(4, 5, 3).astype(np.float32)

# Target size for padding (larger than image dimensions)
targetSize = [8, 10]

# Background color for padding
backgroundColor = 0.5

ARGS = {
    "image": image,
    "targetSize": targetSize,
    "backgroundColor": backgroundColor,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (8, 10, 3) — shape (8, 10, 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
softDice Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/bratsUtils.py
code served (permissive licence) · get_code("3b9462f9056ee95e")
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

torch.manual_seed(0)

# Create tiny tensors for pred and target
# The function uses operations: .sum(dim=(1,2,3)), .float(), > 0.5
# So we need tensors with at least 4 dimensions
# Shape: (batch=2, channels=3, height=4, width=5)
pred = torch.rand(2, 3, 4, 5)
target = torch.randint(0, 2, (2, 3, 4, 5)).float()

ARGS = {
    "pred": pred,
    "target": target,
    "smoothing": 1,
    "nonSquared": 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 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
test_train_val_split Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/brats18_data_loader.py
code served (permissive licence) · get_code("407af86815254b44")
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)

ARGS = {
    'patient_id': 5
}
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
test_train_val_split Ran
ran on a synthesized input
RobinBruegger/PartiallyReversibleUnet/dataProcessing/brats18_validation_data_loader.py
code served (permissive licence) · get_code("ea35d2f3d8516236")
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)

ARGS = {
    'patient_id': 'patient_001'
}
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

Also present in other papers' code

Of 15 samples lifted from this paper's linked repositories, 15 carry a function identity and 0 do not. Of the 15 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 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.

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

One of the key drawbacks of 3D convolutional neural networks for segmentation is their memory footprint, which necessitates compromises in the network architecture in order to fit into a given memory budget. Motivated by the RevNet for image classification, we propose a partially reversible U-Net architecture that reduces memory consumption substantially. The reversible architecture allows us to exactly recover each layer's outputs from the subsequent layer's ones, eliminating the need to store activations for backpropagation. This alleviates the biggest memory bottleneck and enables very deep (theoretically infinitely deep) 3D architectures. On the BraTS challenge dataset, we demonstrate substantial memory savings. We further show that the freed memory can be used for processing the whole field-of-view (FOV) instead of patches. Increasing network depth led to higher segmentation accuracy while growing the memory footprint only by a very small fraction, thanks to the partially reversible architecture.

For agents

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

get_harvested_code_for_paper("1906.06148")
get_code_for_paper("1906.06148")
have("1906.06148")

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

curl https://syntology.ai/api/ran/1906.06148.json

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

[![Syntology run record](https://syntology.ai/api/ran/1906.06148.svg)](https://syntology.ai/paper/1906.06148)

Connect an agent — have() is free.