SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1910.00780 · 2019

How does topology influence gradient propagation and model performance of deep networks with DenseNet-type skip connections?

arXiv · PDF · Open in the Atlas

Code that ran

4 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 1 violated it and 0 have no contract verdict; 2 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/1910.00780.json · badge: /api/ran/1910.00780.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
SLDGroup/NN_Mass canonical (classifier label)
classifier basis: classifier_2025
4 ran · 0 no recorded run
FunctionStatusWhere it lives
accuracy Ran
ran, violated the contract we drafted
SLDGroup/NN_Mass/train_mlp.py
code served (permissive licence) · get_code("1d9cfaf4282b4c13")
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 torch.nn as nn

torch.manual_seed(0)

# The function expects:
# - output: a tensor with .detach(), .cpu(), and .max(1) methods
#   .max(1) returns (values, indices) where indices are predictions
# - target: a tensor with .detach(), .cpu(), and .eq() methods
#   should be 1D with class labels

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

# Create target tensor: batch_size=2, with class indices (0-3)
target = torch.tensor([0, 2])

ARGS = {
    "output": output,
    "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 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
test Ran
ran, honoured the contract we drafted
SLDGroup/NN_Mass/train_cifar.py
code served (permissive licence) · get_code("4800dd5ec9240018")
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. 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
import torch.nn.functional as F
from torch.autograd import Variable
import tempfile

torch.manual_seed(0)

# Small model for testing
class SimpleModel(nn.Module):
    def __init__(self):
        super(SimpleModel, self).__init__()
        self.fc = nn.Linear(4, 3)
        self.nn_mass = 0.5  # Required attribute used in the function
    
    def forward(self, x):
        return self.fc(x)

# Create test data loader (list of batches)
test_loader = [
    (torch.randn(2, 4), torch.randint(0, 3, (2,))),
    (torch.randn(2, 4), torch.randint(0, 3, (2,)))
]

# Create model
model = SimpleModel()

# Create criterion
criterion = nn.CrossEntropyLoss()

# Create temporary file for logging
temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.txt')
file_name = temp_file.name
temp_file.close()

# Build arguments dictionary
ARGS = {
    'epoch': 1,
    'test_loader': test_loader,
    'model': model,
    'file_name': file_name,
    'criterion': criterion,
    'device': torch.device('cpu')
}
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 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
SLDGroup/NN_Mass/train_mlp.py
code served (permissive licence) · get_code("c0bd5cd7c711d4d1")
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
from torch.autograd import Variable

torch.manual_seed(0)

# Define a simple model that takes flattened input and outputs class probabilities
class SimpleModel(nn.Module):
    def __init__(self, input_size, num_classes):
        super(SimpleModel, self).__init__()
        self.fc = nn.Linear(input_size, num_classes)
    
    def forward(self, x):
        return self.fc(x)

# Set up parameters
batch_size = 2
image_height = 4
image_width = 4
num_classes = 3
input_size = image_height * image_width

# Create model
model = SimpleModel(input_size, num_classes)

# Create tiny tensors
# images: batch_size x channels x height x width (2, 1, 4, 4)
images = torch.randn(batch_size, 1, image_height, image_width)

# labels: batch_size (2 class indices)
labels = torch.tensor([0, 2], dtype=torch.long)

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

# Create loss function
criterion = nn.CrossEntropyLoss()

# Device
device = 'cpu'

ARGS = {
    'images': images,
    'labels': labels,
    'optimizer': optimizer,
    'model': model,
    'criterion': criterion,
    'device': device
}
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
train Ran
ran; our contract draft was wrong, not the code
SLDGroup/NN_Mass/train_cifar.py
code served (permissive licence) · get_code("35d00fe15613b708")
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
from torch.autograd import Variable

torch.manual_seed(0)

# Define a simple model
class SimpleModel(nn.Module):
    def __init__(self):
        super(SimpleModel, self).__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create tiny tensors (batch_size=2, input_dim=4, output_dim=3)
images = torch.randn(2, 4)
labels = torch.randint(0, 3, (2,))

# Create optimizer
model = SimpleModel()
optimizer = torch.optim.SGD(model.parameters(), lr=0.01)

# Create loss criterion
criterion = nn.CrossEntropyLoss()

# Set device to CPU
device = 'cpu'

ARGS = {
    'images': images,
    'labels': labels,
    'optimizer': optimizer,
    'model': model,
    'criterion': criterion,
    'device': device
}
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

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 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 4 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

DenseNets introduce concatenation-type skip connections that achieve state-of-the-art accuracy in several computer vision tasks. In this paper, we reveal that the topology of the concatenation-type skip connections is closely related to the gradient propagation which, in turn, enables a predictable behavior of DNNs' test performance. To this end, we introduce a new metric called NN-Mass to quantify how effectively information flows through DNNs. Moreover, we empirically show that NN-Mass also works for other types of skip connections, e.g., for ResNets, Wide-ResNets (WRNs), and MobileNets, which contain addition-type skip connections (i.e., residuals or inverted residuals). As such, for both DenseNet-like CNNs and ResNets/WRNs/MobileNets, our theoretically grounded NN-Mass can identify models with similar accuracy, despite having significantly different size/compute requirements. Detailed experiments on both synthetic and real datasets (e.g., MNIST, CIFAR-10, CIFAR-100, ImageNet) provide extensive evidence for our insights. Finally, the closed-form equation of our NN-Mass enables us to design significantly compressed DenseNets (for CIFAR-10) and MobileNets (for ImageNet) directly at initialization without time-consuming training and/or searching.

For agents

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

get_harvested_code_for_paper("1910.00780")
get_code_for_paper("1910.00780")
have("1910.00780")

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

curl https://syntology.ai/api/ran/1910.00780.json

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

[![Syntology run record](https://syntology.ai/api/ran/1910.00780.svg)](https://syntology.ai/paper/1910.00780)

Connect an agent — have() is free.