SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2001.01431 · 2020

Deeper Insights into Weight Sharing in Neural Architecture Search

arXiv · PDF · Open in the Atlas

Code that ran

4 of 6 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 2 have no contract verdict; 2 failures of our instrument; 2 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/2001.01431.json · badge: /api/ran/2001.01431.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
ultmaster/deeper-insights-weight-sharing canonical (classifier label)
classifier basis: census
4 ran · 2 no recorded run
FunctionStatusWhere it lives
config_uglify Ran
ran on a synthesized input
ultmaster/deeper-insights-weight-sharing/submit_utils.py
code served (permissive licence) · get_code("fcde1a98c0385d78")
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 os
import random
import string
import tempfile

# The function uses the 'experiment' parameter as a file path
# It calls os.path.relpath, os.path.splitext on it
# So we need to provide a valid file path string

# Create a temporary directory structure to simulate experiment paths
temp_dir = tempfile.gettempdir()
experiments_dir = os.path.join(temp_dir, 'experiments')
os.makedirs(experiments_dir, exist_ok=True)

# Create a nested experiment path
nested_exp_dir = os.path.join(experiments_dir, 'subdir', 'my_experiment.txt')
os.makedirs(os.path.dirname(nested_exp_dir), exist_ok=True)

# Ensure the file exists or at least the path is valid
experiment_path = nested_exp_dir

# Change to temp directory so relpath works correctly from 'experiments'
original_cwd = os.getcwd()
os.chdir(temp_dir)

ARGS = {
    'experiment': experiment_path
}

os.chdir(original_cwd)
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
data_transforms Ran
ran; our contract draft was wrong, not the code
ultmaster/deeper-insights-weight-sharing/preproc.py
code served (permissive licence) · get_code("1d6c0c0dff1c1274")
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 numpy as np
import torchvision.transforms as transforms

torch.manual_seed(0)
np.random.seed(0)

ARGS = {
    'dataset': 'cifar10',
    'cutout_length': 8
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, torchvision — 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
get_parser Ran
ran; our contract draft was wrong, not the code
ultmaster/deeper-insights-weight-sharing/config.py
pointer only (licence: MIT); the stored copy get_code serves, from harveyp123/maxk-gnn, is pointer only (licence: NONE) · get_code("bf0bc305c2302fec")
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 harveyp123/maxk-gnn, 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.

parse Ran
ran on a synthesized input
ultmaster/deeper-insights-weight-sharing/genotypes.py
code served (permissive licence) · get_code("7ac97563dbd4f579")
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 3, 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

torch.manual_seed(0)

# alpha is a ParameterList with multiple Parameter tensors
# Each Parameter has shape (n_edges, n_ops)
# The function iterates over alpha, so we need multiple edge tensors
# For each edge tensor, topk is called with k=1 first, then k=2 on flattened max values

# Create a ParameterList with 3 edge tensors
# n_edges=4 (number of edges per node), n_ops=4 (number of primitives)
alpha = torch.nn.ParameterList([
    torch.nn.Parameter(torch.randn(4, 4)),  # Node 1: 4 edges, 4 operations
    torch.nn.Parameter(torch.randn(4, 4)),  # Node 2: 4 edges, 4 operations
    torch.nn.Parameter(torch.randn(4, 4)),  # Node 3: 4 edges, 4 operations
])

# k is the number of top edges to select per node
k = 2

ARGS = {
    'alpha': alpha,
    '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 list, ndim 3, 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
drop_path_ Not yet run
unverified (harvested, no recorded run)
ultmaster/deeper-insights-weight-sharing/models/ops.py
code served (permissive licence) · get_code("654a527a9f7f0bc9")
How it ran

no invocation recorded.

from_str Not yet run
unverified (harvested, no recorded run)
ultmaster/deeper-insights-weight-sharing/genotypes.py
code served (permissive licence) · get_code("eb9857d523d512dc")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 4 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 1 later) and 2 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
drop_path_ 2 other papers
2 dated earlier
1805.09501 (2018-05)
AutoAugment: Learning Augmentation Policies from Data
1806.09055 (2018-06)
DARTS: Differentiable Architecture Search
from_str 2 other papers
2 dated earlier
1805.09501 (2018-05)
AutoAugment: Learning Augmentation Policies from Data
1806.09055 (2018-06)
DARTS: Differentiable Architecture Search
get_parser 2 other papers
1 dated earlier, 1 later
1806.09055 (2018-06)
DARTS: Differentiable Architecture Search
2312.08656 (2023-12)
MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training
data_transforms 1 other paper
1 dated earlier
1806.09055 (2018-06)
DARTS: Differentiable Architecture Search

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

With the success of deep neural networks, Neural Architecture Search (NAS) as a way of automatic model design has attracted wide attention. As training every child model from scratch is very time-consuming, recent works leverage weight-sharing to speed up the model evaluation procedure. These approaches greatly reduce computation by maintaining a single copy of weights on the super-net and share the weights among every child model. However, weight-sharing has no theoretical guarantee and its impact has not been well studied before. In this paper, we conduct comprehensive experiments to reveal the impact of weight-sharing: (1) The best-performing models from different runs or even from consecutive epochs within the same run have significant variance; (2) Even with high variance, we can extract valuable information from training the super-net with shared weights; (3) The interference between child models is a main factor that induces high variance; (4) Properly reducing the degree of weight sharing could effectively reduce variance and improve performance.

For agents

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

get_harvested_code_for_paper("2001.01431")
get_code_for_paper("2001.01431")
have("2001.01431")

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

curl https://syntology.ai/api/ran/2001.01431.json

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

[![Syntology run record](https://syntology.ai/api/ran/2001.01431.svg)](https://syntology.ai/paper/2001.01431)

Connect an agent — have() is free.