SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1810.05331 · 2018

Dynamic Channel Pruning: Feature Boosting and Suppression

arXiv · PDF · Open in the Atlas

Code that ran

5 of 7 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 3 have no contract verdict; 2 failures of our instrument; 2 no recorded run), measured 2026-11-21; 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/1810.05331.json · badge: /api/ran/1810.05331.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
deep-fry/mayo canonical (classifier label)
classifier basis: census
3 ran · 0 no recorded run
YOUSIKI/PyTorch-FBS reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 2 no recorded run
FunctionStatusWhere it lives
get_in_channels Ran
ran; our contract draft was wrong, not the code
YOUSIKI/PyTorch-FBS/models/custom.py
code served (permissive licence) · get_code("c61b8516007e275a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned 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)

# Test case 1: Sequential with Linear inside
m1 = nn.Sequential(
    nn.Linear(10, 5),
    nn.ReLU()
)

# Test case 2: Direct Linear
m2 = nn.Linear(8, 4)

# Test case 3: Conv2d
m3 = nn.Conv2d(3, 16, kernel_size=3)

# Test case 4: BatchNorm1d
m4 = nn.BatchNorm1d(32)

# Test case 5: Conv1d
m5 = nn.Conv1d(2, 8, kernel_size=5)

# Test case 6: Conv3d
m6 = nn.Conv3d(1, 4, kernel_size=3)

# Test case 7: BatchNorm2d
m7 = nn.BatchNorm2d(16)

# Test case 8: BatchNorm3d
m8 = nn.BatchNorm3d(8)

# Test case 9: Sequential with Conv2d
m9 = nn.Sequential(
    nn.Conv2d(3, 32, kernel_size=3),
    nn.ReLU(),
    nn.MaxPool2d(2)
)

# Test case 10: Sequential with BatchNorm
m10 = nn.Sequential(
    nn.BatchNorm1d(64),
    nn.ReLU()
)

# Test case 11: Module that returns None (e.g., ReLU)
m11 = nn.ReLU()

# Test case 12: Nested Sequential
m12 = nn.Sequential(
    nn.Sequential(
        nn.Linear(16, 8)
    ),
    nn.ReLU()
)

ARGS = {
    "m": m1
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
get_out_channels Ran
ran; our contract draft was wrong, not the code
YOUSIKI/PyTorch-FBS/models/custom.py
code served (permissive licence) · get_code("8b74cd9e7bbcae25")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned 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
from typing import Optional

torch.manual_seed(0)

# Test case 1: Linear module
linear_module = nn.Linear(10, 20)

# Test case 2: Conv2d module
conv2d_module = nn.Conv2d(3, 16, kernel_size=3)

# Test case 3: BatchNorm2d module
batchnorm_module = nn.BatchNorm2d(32)

# Test case 4: Sequential with Linear
sequential_linear = nn.Sequential(
    nn.Linear(5, 10),
    nn.ReLU(),
    nn.Linear(10, 15)
)

# Test case 5: Sequential with Conv2d
sequential_conv = nn.Sequential(
    nn.Conv2d(3, 8, kernel_size=3),
    nn.ReLU(),
    nn.Conv2d(8, 16, kernel_size=3)
)

# Test case 6: Sequential with BatchNorm
sequential_batchnorm = nn.Sequential(
    nn.BatchNorm1d(64),
    nn.ReLU()
)

# Test case 7: Nested Sequential
nested_sequential = nn.Sequential(
    nn.Sequential(
        nn.Linear(4, 8)
    ),
    nn.ReLU(),
    nn.Linear(8, 12)
)

# Test case 8: Module that returns None
class CustomModule(nn.Module):
    def forward(self, x):
        return x

custom_module = CustomModule()

# Test case 9: Conv1d module
conv1d_module = nn.Conv1d(2, 4, kernel_size=3)

# Test case 10: Conv3d module
conv3d_module = nn.Conv3d(1, 8, kernel_size=3)

ARGS = {
    "m": linear_module
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
mask_join Ran
ran on a synthesized input
deep-fry/mayo/mayo/net/tf/estimate.py
code served (permissive licence) · get_code("3088ca7dabe32a62")
How it ran

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

import functools
import operator

torch = None  # not needed for this function

# mask_join takes two parameters:
# 1. masks: an iterable of items that are either lists or non-list objects
# 2. reducer: a callable that reduces a sequence of items to a single item

# The function:
# - Finds the max length among list items in masks
# - Broadcasts non-list items to that length by repeating them
# - Zips corresponding elements across all masks
# - Applies reducer to each group

# Create test data
masks = [
    [1, 2, 3],           # list of length 3
    5,                   # non-list, will be broadcast to [5, 5, 5]
    [10, 20, 30]         # list of length 3
]

# reducer should be a function that takes an iterable and returns a single value
# operator.add is a simple reducer that works with reduce
reducer = operator.add

ARGS = {
    "masks": masks,
    "reducer": reducer
}
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 int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
multiply Ran
ran on a synthesized input
deep-fry/mayo/mayo/net/tf/estimate.py
code served (permissive licence) · get_code("0667f26e9f8bd466")
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

torch.manual_seed(0)

# The function multiply() takes 'items' and multiplies all elements together
# items should be an iterable of numeric values
items = [2, 3, 4, 5]

ARGS = {
    "items": items
}
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
use_name_not_scope Ran
ran on a synthesized input
deep-fry/mayo/mayo/net/tf/transform.py
code served (permissive licence) · get_code("9a0f792b34207b94")
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 dict. 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)

ARGS = {
    'params': {'scope': 'test_scope', 'other_key': 'other_value'}
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
conv1x1 Not yet run
unverified (harvested, no recorded run)
YOUSIKI/PyTorch-FBS/models/resnet.py
code served (permissive licence) · get_code("350baadaa22ae4dd")
How it ran

no invocation recorded.

conv3x3 Not yet run
unverified (harvested, no recorded run)
YOUSIKI/PyTorch-FBS/models/resnet.py
code served (permissive licence) · get_code("aa53dd1441e0a7d8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 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 7 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 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Making deep convolutional neural networks more accurate typically comes at the cost of increased computational and memory resources. In this paper, we reduce this cost by exploiting the fact that the importance of features computed by convolutional layers is highly input-dependent, and propose feature boosting and suppression (FBS), a new method to predictively amplify salient convolutional channels and skip unimportant ones at run-time. FBS introduces small auxiliary connections to existing convolutional layers. In contrast to channel pruning methods which permanently remove channels, it preserves the full network structures and accelerates convolution by dynamically skipping unimportant input and output channels. FBS-augmented networks are trained with conventional stochastic gradient descent, making it readily available for many state-of-the-art CNNs. We compare FBS to a range of existing channel pruning and dynamic execution schemes and demonstrate large improvements on ImageNet classification. Experiments show that FBS can respectively provide $5\times$ and $2\times$ savings in compute on VGG-16 and ResNet-18, both with less than $0.6\%$ top-5 accuracy loss.

For agents

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

get_harvested_code_for_paper("1810.05331")
get_code_for_paper("1810.05331")
have("1810.05331")

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

curl https://syntology.ai/api/ran/1810.05331.json

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

[![Syntology run record](https://syntology.ai/api/ran/1810.05331.svg)](https://syntology.ai/paper/1810.05331)

Connect an agent — have() is free.