SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2108.06822 · 2021

CONet: Channel Optimization for Convolutional Neural Networks

arXiv · PDF · Open in the Atlas

Code that ran

7 of 10 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 4 have no contract verdict; 3 failures of our instrument; 3 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/2108.06822.json · badge: /api/ran/2108.06822.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
mahdihosseini/conet canonical (classifier label)
classifier basis: census
7 ran · 3 no recorded run
FunctionStatusWhere it lives
EVBMF Ran
ran; our contract draft was wrong, not the code
mahdihosseini/conet/CONet/VBMF.py
code served (permissive licence) · get_code("0493f54aeb087086")
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 numpy as np

torch.manual_seed(0)

# Create a small test matrix Y with shape (L, M) where L <= M
# Using L=3, M=5 to have distinct dimensions
L, M = 3, 5
Y = torch.randn(L, M, dtype=torch.float32)

# Optional parameters
sigma2 = None  # Will be estimated if None
H = None  # Will default to L if None

ARGS = {
    "Y": Y,
    "sigma2": sigma2,
    "H": H,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, 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
VBMF Ran
ran; our contract draft was wrong, not the code
mahdihosseini/conet/CONet/VBMF.py
code served (permissive licence) · get_code("0b42d960ea81c183")
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 test matrix Y with shape (L, M) where L <= M
np.random.seed(0)
L, M = 4, 6
Y = np.random.randn(L, M)

# cacb is the product of prior variances - a scalar
cacb = 1.0

# sigma2 can be None (will be estimated) or a float
sigma2 = None

# H is the maximum rank, can be None (will default to L) or an integer
H = None

ARGS = {
    'Y': Y,
    'cacb': cacb,
    'sigma2': sigma2,
    'H': H
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
VBsigma2 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
mahdihosseini/conet/CONet/VBMF.py
code served (permissive licence) · get_code("9e8ce2d162e22cdd")
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

# Analyze the function to understand parameter requirements:
# - sigma2: scalar float (noise variance)
# - L: int (appears in calculations)
# - M: int (appears in calculations)
# - cacb: scalar float (appears in divisions and multiplications)
# - s: 1D numpy array of singular values (sorted in descending order)
# - residual: scalar float (residual sum of squares)

# Create small test data
np.random.seed(0)

# s should be a 1D array of singular values, typically sorted descending
s = np.array([5.0, 3.0, 1.5, 0.5])

# Scalar parameters
sigma2 = 1.0
L = 3
M = 4
cacb = 2.0
residual = 2.5

ARGS = {
    "sigma2": sigma2,
    "L": L,
    "M": M,
    "cacb": cacb,
    "s": s,
    "residual": residual
}
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
compile_adaptive_files Ran
ran on a synthesized input
mahdihosseini/conet/CONet/adaptive_graph.py
code served (permissive licence) · get_code("721a4b3213ecc252")
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.

# Setup code for compile_adaptive_files

ARGS = {
    "file_name": "path/trial5_data.txt",
    "num_trials": 3
}
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
find_cell_first_conv_idx Ran
ran on a synthesized input
mahdihosseini/conet/CONet/darts_scaling_algorithms.py
code served (permissive licence) · get_code("025c9771197fa64f")
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 copy

def find_cell_first_conv_idx(cell_list):
    normal_cell_layers = 24
    reduction_cell_layers = 2
    red_normal_cell_layers = 25
    cur_layer_count = 0
    reduction_cell_idx = [(len(cell_list) - 1) // 3 + 1, (len(cell_list) - 1) * 2 // 3 + 1]
    prev_cell_is_reduction = False
    cells_first_layer_idx = []
    for cur_cell_idx, cur_cell in enumerate(cell_list):
        if cur_cell_idx == 0:
            cur_layer_count = 0
        elif cur_cell_idx == 1:
            cur_layer_count = 1
        elif cur_cell_idx - 1 in reduction_cell_idx:
            prev_cell_is_reduction = True
            cur_layer_count += reduction_cell_layers
        elif prev_cell_is_reduction:
            cur_layer_count += red_normal_cell_layers
            prev_cell_is_reduction = False
        else:
            cur_layer_count += normal_cell_layers
        cells_first_layer_idx.append(cur_layer_count)
    return copy.deepcopy(cells_first_layer_idx)


class DummyCell:
    """Minimal stand-in for a cell object"""
    def __init__(self, name):
        self.name = name


ARGS = {
    "cell_list": [
        DummyCell("cell_0"),
        DummyCell("cell_1"),
        DummyCell("cell_2"),
        DummyCell("cell_3"),
        DummyCell("cell_4"),
        DummyCell("cell_5"),
        DummyCell("cell_6"),
    ]
}
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
find_cell_first_conv_idx Ran
ran on a synthesized input
mahdihosseini/conet/CONet/dartsplus_scaling_algorithms.py
code served (permissive licence) · get_code("7adbbc7cf50d72ed")
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 copy

def find_cell_first_conv_idx(cell_list):
    normal_cell_layers = 26
    reduction_cell_layers = 8
    red_normal_cell_layers = 27
    cur_layer_count = 0
    reduction_cell_idx = [(len(cell_list) - 1) // 3 + 1, (len(cell_list) - 1) * 2 // 3 + 1]
    prev_cell_is_reduction = False
    cells_first_layer_idx = []
    for cur_cell_idx, cur_cell in enumerate(cell_list):
        if cur_cell_idx == 0:
            cur_layer_count = 0
        elif cur_cell_idx == 1:
            cur_layer_count = 1
        elif cur_cell_idx - 1 in reduction_cell_idx:
            prev_cell_is_reduction = True
            cur_layer_count += reduction_cell_layers
        elif prev_cell_is_reduction:
            cur_layer_count += red_normal_cell_layers
            prev_cell_is_reduction = False
        else:
            cur_layer_count += normal_cell_layers
        cells_first_layer_idx.append(cur_layer_count)
    return copy.deepcopy(cells_first_layer_idx)


class StandInCell:
    """Minimal stand-in cell object for the function."""
    def __init__(self, name):
        self.name = name


cell_list = [
    StandInCell("cell_0"),
    StandInCell("cell_1"),
    StandInCell("cell_2"),
    StandInCell("cell_3"),
    StandInCell("cell_4"),
    StandInCell("cell_5"),
    StandInCell("cell_6"),
]

ARGS = {
    "cell_list": cell_list,
}
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
get_cells_type Ran
ran on a synthesized input
mahdihosseini/conet/CONet/darts_scaling_algorithms.py
code served (permissive licence) · get_code("2668c073a774ff76")
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 copy

# Analyze the function: get_cells_type takes a cell_list parameter
# The function iterates through cell_list with enumerate, so it needs to be iterable
# It accesses len(cell_list), so it needs to support len()
# It checks cur_cell_idx against integers and uses cur_cell in the loop but doesn't actually use cur_cell's properties
# So cell_list just needs to be a list of items (can be any objects)

# Create a minimal cell_list - the function doesn't actually use properties of individual cells,
# it only iterates and checks indices
cell_list = [
    {'name': 'cell_0'},
    {'name': 'cell_1'},
    {'name': 'cell_2'},
    {'name': 'cell_3'},
    {'name': 'cell_4'},
    {'name': 'cell_5'},
]

ARGS = {
    'cell_list': cell_list,
}
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 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_layer_conv_ranks Not yet run
unverified (harvested, no recorded run)
mahdihosseini/conet/CONet/darts_scaling_algorithms.py
code served (permissive licence) · get_code("3bf1f494c6a5582c")
How it ran

no invocation recorded.

load_meta_file Not yet run
unverified (harvested, no recorded run)
mahdihosseini/conet/CONet/datasets.py
code served (permissive licence) · get_code("63a8dc3bb02af22f")
How it ran

no invocation recorded.

parse_devkit_archive Not yet run
unverified (harvested, no recorded run)
mahdihosseini/conet/CONet/datasets.py
code served (permissive licence) · get_code("1dcb024fb2f715dd")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 distinct functions with an identity, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 2 later) and 5 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
VBMF 2 other papers
1 dated earlier, 1 later
1412.6980 (2014-12)
Adam: A Method for Stochastic Optimization
2111.14056 (2021-11)
Towards Robust and Automatic Hyper-Parameter Tunning
VBsigma2 2 other papers
1 dated earlier, 1 later
1412.6980 (2014-12)
Adam: A Method for Stochastic Optimization
2111.14056 (2021-11)
Towards Robust and Automatic Hyper-Parameter Tunning
EVBMF 1 other paper
1 dated earlier
1412.6980 (2014-12)
Adam: A Method for Stochastic Optimization
load_meta_file 1 other paper
1 dated earlier
1412.6980 (2014-12)
Adam: A Method for Stochastic Optimization
parse_devkit_archive 1 other paper
1 dated earlier
1412.6980 (2014-12)
Adam: A Method for Stochastic Optimization

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

Neural Architecture Search (NAS) has shifted network design from using human intuition to leveraging search algorithms guided by evaluation metrics. We study channel size optimization in convolutional neural networks (CNN) and identify the role it plays in model accuracy and complexity. Current channel size selection methods are generally limited by discrete sample spaces while suffering from manual iteration and simple heuristics. To solve this, we introduce an efficient dynamic scaling algorithm -- CONet -- that automatically optimizes channel sizes across network layers for a given CNN. Two metrics -- "\textit{Rank}" and "\textit{Rank Average Slope}" -- are introduced to identify the information accumulated in training. The algorithm dynamically scales channel sizes up or down over a fixed searching phase. We conduct experiments on CIFAR10/100 and ImageNet datasets and show that CONet can find efficient and accurate architectures searched in ResNet, DARTS, and DARTS+ spaces that outperform their baseline models. This document supersedes previously published paper in ICCV2021-NeurArch workshop. An additional section is included on manual scaling of channel size in CNNs to numerically validate of the metrics used in searching optimum channel configurations in CNNs.

For agents

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

get_harvested_code_for_paper("2108.06822")
get_code_for_paper("2108.06822")
have("2108.06822")

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

curl https://syntology.ai/api/ran/2108.06822.json

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

[![Syntology run record](https://syntology.ai/api/ran/2108.06822.svg)](https://syntology.ai/paper/2108.06822)

Connect an agent — have() is free.