SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2009.01027 · 2020

DARTS-: Robustly Stepping out of Performance Collapse Without Indicators

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 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 6 have no contract verdict; 0 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/2009.01027.json · badge: /api/ran/2009.01027.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
Meituan-AutoML/DARTS- canonical (classifier label)
classifier basis: census
6 ran · 2 no recorded run
FunctionStatusWhere it lives
gen_beta Ran
ran on a synthesized input
Meituan-AutoML/DARTS-/plots/plot_beta_vs_valid_loss.py
code served (permissive licence) · get_code("a48bdfe2a6bf0357")
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

torch.manual_seed(0)

ARGS = {
    'beta_decay': 'linear',
    'beta': 1.0
}
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_moving_average_3 Ran
ran on a synthesized input
Meituan-AutoML/DARTS-/plots/eigen_dist.py
code served (permissive licence) · get_code("30bb0400a2dd1188")
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 numpy as np

# Analyze the function: it takes 'eigens' which is indexed with [start:stop] syntax
# and passed to np.mean(), so it should be array-like (list or numpy array)
# The function accesses eigens[0:2], eigens[-2:], eigens[start_index-1:start_index+2]
# and checks len(eigens), so it needs to support slicing and len()

# Create a simple numpy array of eigenvalues
eigens = np.array([1.5, 2.3, 3.1, 4.2, 5.0, 6.7])

ARGS = {
    "eigens": eigens
}
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 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_moving_average_5 Ran
ran on a synthesized input
Meituan-AutoML/DARTS-/plots/eigen_dist.py
code served (permissive licence) · get_code("db90642e32525cb0")
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 numpy as np

# Analyze the function: get_moving_average_5 takes eigens as parameter
# It expects eigens to be indexable (supports slicing) and have len()
# It passes slices to np.mean(), so it should be array-like with numeric values
# The function computes a moving average over windows of size 3-5

# Create a simple numpy array with numeric values
eigens = np.array([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0])

ARGS = {
    "eigens": eigens
}
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 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
read_json Ran
ran on a synthesized input
Meituan-AutoML/DARTS-/plots/plot_beta_vs_valid_loss.py
code served (permissive licence) · get_code("35394520b14b68c3")
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 json
import tempfile
import os

# Create a temporary JSON lines file with valid content
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test.jsonl")

with open(temp_file, 'w') as f:
    f.write('{"key1": "value1", "key2": 42}\n')
    f.write('{"key1": "value2", "key2": 43}\n')

ARGS = {
    "file": temp_file
}
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 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
singleton Ran
ran on a synthesized input
Meituan-AutoML/DARTS-/src/utils.py
code served (permissive licence) · get_code("de6f3cc2917b53d0")
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 function. 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 functools
import torch

class DummyClass:
    """A simple class to use as the cls argument for singleton decorator"""
    def __init__(self, value=42):
        self.value = value
    
    def get_value(self):
        return self.value

torch.manual_seed(0)

ARGS = {
    'cls': DummyClass,
    'args': (),
    'kw': {}
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
stop_criteria Ran
ran on a synthesized input
Meituan-AutoML/DARTS-/plots/eigen_dist.py
code served (permissive licence) · get_code("101678ccfacaef06")
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 tempfile
import os

# Create a temporary log file with genotype data
temp_log = tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.log')
log_file_path = temp_log.name

# Write test log entries with genotype data
# We need at least 2*13 = 26 lines with genotype (to access lines[2*i] where i goes up to ~12)
genotypes = [f"genotype_{j}" for j in range(30)]
for i, gen in enumerate(genotypes):
    temp_log.write(f"step {i}: genotype = {repr(gen)}\n")
temp_log.close()

# Create moving_averages list
# We need len(moving_averages) > 9 (since loop goes range(5, len-4))
# We want to trigger the condition at some point: moving_averages[i+4] / moving_averages[i] > factor
# Let's create a list where this happens around i=5
moving_averages = [100.0, 100.0, 100.0, 100.0, 100.0, 100.0, 150.0, 150.0, 150.0, 150.0, 10.0]
# At i=5: moving_averages[9] / moving_averages[5] = 150.0 / 100.0 = 1.5 > 1.3 (default factor)

ARGS = {
    "moving_averages": moving_averages,
    "log_file": log_file_path,
    "factor": 1.3
}

# Cleanup will happen after function execution
import atexit
atexit.register(lambda: os.unlink(log_file_path))
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 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
accuracy Not yet run
unverified (harvested, no recorded run)
Meituan-AutoML/DARTS-/src/utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from yuxwind/exactcompression, is pointer only (licence: NONE) · get_code("469bf48752905889")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from yuxwind/exactcompression, 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. No invocation recorded.

count_parameters_in_MB Not yet run
unverified (harvested, no recorded run)
Meituan-AutoML/DARTS-/src/utils.py
code served (permissive licence) · get_code("3e816289579cc9e9")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 1 the same month, 2 later) and 3 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
accuracy 71 other papers
40 dated earlier, 1 the same month, 30 later · 2 only as an assembly brick
1505.05424 (2015-05)
Weight Uncertainty in Neural Networks
1705.07204 (2017-05)
Ensemble Adversarial Training: Attacks and Defenses
1802.10399 (2018-02)
Compressing Neural Networks using the Variational Information Bottleneck
and 68 more not named here
count_parameters_in_MB 8 other papers
6 dated earlier, 2 later
1806.09055 (2018-06)
DARTS: Differentiable Architecture Search
1912.00195 (2019-12)
SGAS: Sequential Greedy Architecture Search
2004.14557 (2020-04)
Learning Deformable Image Registration from Optimization: Perspective, Modules, Bilevel Training and Beyond
and 5 more not named here
get_moving_average_3 1 other paper
1 dated earlier
1909.09656 (2019-09)
Understanding and Robustifying Differentiable Architecture Search
get_moving_average_5 1 other paper
1 dated earlier
1909.09656 (2019-09)
Understanding and Robustifying Differentiable Architecture Search
stop_criteria 1 other paper
1 dated earlier
1909.09656 (2019-09)
Understanding and Robustifying 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

Despite the fast development of differentiable architecture search (DARTS), it suffers from long-standing performance instability, which extremely limits its application. Existing robustifying methods draw clues from the resulting deteriorated behavior instead of finding out its causing factor. Various indicators such as Hessian eigenvalues are proposed as a signal to stop searching before the performance collapses. However, these indicator-based methods tend to easily reject good architectures if the thresholds are inappropriately set, let alone the searching is intrinsically noisy. In this paper, we undertake a more subtle and direct approach to resolve the collapse. We first demonstrate that skip connections have a clear advantage over other candidate operations, where it can easily recover from a disadvantageous state and become dominant. We conjecture that this privilege is causing degenerated performance. Therefore, we propose to factor out this benefit with an auxiliary skip connection, ensuring a fairer competition for all operations. We call this approach DARTS-. Extensive experiments on various datasets verify that it can substantially improve robustness. Our code is available at https://github.com/Meituan-AutoML/DARTS- .

For agents

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

get_harvested_code_for_paper("2009.01027")
get_code_for_paper("2009.01027")
have("2009.01027")

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

curl https://syntology.ai/api/ran/2009.01027.json

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

[![Syntology run record](https://syntology.ai/api/ran/2009.01027.svg)](https://syntology.ai/paper/2009.01027)

Connect an agent — have() is free.