SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1806.10282 · 2018

Auto-Keras: An Efficient Neural Architecture Search System

arXiv · PDF · Open in the Atlas

Code that ran

6 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; 1 honoured the contract we drafted, 0 violated it and 3 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/1806.10282.json · badge: /api/ran/1806.10282.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
keras-team/autokeras community (classifier label)
classifier basis not recorded on the link
3 ran · 0 no recorded run
finefun0/autokeras reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
FunctionStatusWhere it lives
get_global_average_pooling Ran
ran on a synthesized input
keras-team/autokeras/autokeras/utils/layer_utils.py
code served (permissive licence) · get_code("b8da594ade88788a")
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 type. 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
from keras import layers

# The function takes a 'shape' parameter and uses len(shape) - 3 as an index
# into a list of GlobalAveragePooling layer classes.
# 
# For len(shape) - 3 to be valid:
# - If shape has length 3, index is 0 -> GlobalAveragePooling1D
# - If shape has length 4, index is 1 -> GlobalAveragePooling2D
# - If shape has length 5, index is 2 -> GlobalAveragePooling3D
#
# We'll test with a 4D shape (for 2D pooling) which is a common case

shape = (2, 3, 4, 5)  # 4D shape: (batch, channels, height, width) -> GlobalAveragePooling2D

ARGS = {
    "shape": shape
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
keras, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
get_global_max_pooling Ran
ran on a synthesized input
keras-team/autokeras/autokeras/utils/layer_utils.py
code served (permissive licence) · get_code("1612a150609ebec9")
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 type. 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
from keras import layers

# The function takes a 'shape' parameter and uses len(shape) - 3 to index into a list
# So shape should be a tuple/list with length 3, 4, or 5 to get valid indices (0, 1, or 2)
# For a 1D pooling layer, we need len(shape) = 3
# For a 2D pooling layer, we need len(shape) = 4
# For a 3D pooling layer, we need len(shape) = 5

# Let's use a shape for 2D pooling (most common case)
# Shape should represent (batch, height, width, channels) or similar tensor dimensions
shape = (2, 4, 8, 3)  # length 4, so will select GlobalMaxPool2D (index 1)

ARGS = {
    "shape": shape
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
keras, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
get_max_pooling Ran
ran on a synthesized input
keras-team/autokeras/autokeras/utils/layer_utils.py
code served (permissive licence) · get_code("ea28a73c015bcee9")
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 type. 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.

from keras import layers

# The function takes a 'shape' parameter and uses len(shape) to index into a list
# of MaxPool layer classes. We need to provide a shape tuple.
# - If len(shape) == 3, it returns MaxPool1D (shape has 3 elements)
# - If len(shape) == 4, it returns MaxPool2D (shape has 4 elements)
# - If len(shape) == 5, it returns MaxPool3D (shape has 5 elements)

# Let's test with a 2D case (4D shape including batch and channels)
shape = (2, 4, 8, 3)  # batch=2, channels=4, height=8, width=3

ARGS = {
    'shape': shape
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
keras — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
layer_distance Ran
ran, honoured the contract we drafted
finefun0/autokeras/autokeras/bayesian.py
code served (permissive licence) · get_code("4d3ba6ed2b564e50")
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 float, ndim 0, dtype family float, 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 layer_distance takes two numeric arguments a and b
# and computes: abs(a - b) * 1.0 / max(a, b)
# These can be scalars, ints, or floats

a = 5
b = 3

ARGS = {
    'a': a,
    'b': b
}
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 float, ndim 0, dtype family float, 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
layers_distance Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
finefun0/autokeras/autokeras/bayesian.py
code served (permissive licence) · get_code("3e2139fba91b7ebf")
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, 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 numpy as np

def layer_distance(a, b):
    return abs(a - b) * 1.0 / max(a, b)

def layers_distance(list_a, list_b):
    len_a = len(list_a)
    len_b = len(list_b)
    f = np.zeros((len_a + 1, len_b + 1))
    f[-1][-1] = 0
    for i in range(-1, len_a):
        f[i][-1] = i + 1
    for j in range(-1, len_b):
        f[-1][j] = j + 1
    for i in range(len_a):
        for j in range(len_b):
            f[i][j] = min(f[i][j - 1] + 1, f[i - 1][j] + 1, f[i - 1][j - 1] + layer_distance(list_a[i], list_b[j]))
    return f[len_a - 1][len_b - 1]

# Construct arguments
list_a = [10, 20, 30, 15]
list_b = [12, 25, 32]

ARGS = {
    "list_a": list_a,
    "list_b": list_b
}
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, 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
skip_connection_distance Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
finefun0/autokeras/autokeras/bayesian.py
code served (permissive licence) · get_code("82283bf125924649")
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 float, ndim 0, dtype family float, 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)

# Analyzing the function:
# - Takes parameters 'a' and 'b'
# - Both are indexed with [0], [1], [2]
# - So they are sequences/tuples with at least 3 elements
# - a[0], a[1], a[2] are accessed
# - b[0], b[1], b[2] are accessed
# - a[0] and a[1] are subtracted (abs(a[1] - a[0]))
# - b[0] and b[1] are subtracted (abs(b[1] - b[0]))
# - a[2] and b[2] are compared for equality
# - All operations are numeric (abs, max, division)

# Create valid arguments as tuples/lists with numeric values
a = (10, 25, 1)  # (start, end, type)
b = (15, 30, 1)  # (start, end, type)

ARGS = {
    "a": a,
    "b": b
}
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 float, ndim 0, dtype family float, 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

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, 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 6 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

Neural architecture search (NAS) has been proposed to automatically tune deep neural networks, but existing search algorithms, e.g., NASNet, PNAS, usually suffer from expensive computational cost. Network morphism, which keeps the functionality of a neural network while changing its neural architecture, could be helpful for NAS by enabling more efficient training during the search. In this paper, we propose a novel framework enabling Bayesian optimization to guide the network morphism for efficient neural architecture search. The framework develops a neural network kernel and a tree-structured acquisition function optimization algorithm to efficiently explores the search space. Intensive experiments on real-world benchmark datasets have been done to demonstrate the superior performance of the developed framework over the state-of-the-art methods. Moreover, we build an open-source AutoML system based on our method, namely Auto-Keras. The system runs in parallel on CPU and GPU, with an adaptive search strategy for different GPU memory limits.

For agents

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

get_harvested_code_for_paper("1806.10282")
get_code_for_paper("1806.10282")
have("1806.10282")

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

curl https://syntology.ai/api/ran/1806.10282.json

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

[![Syntology run record](https://syntology.ai/api/ran/1806.10282.svg)](https://syntology.ai/paper/1806.10282)

Connect an agent — have() is free.