SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1605.07648 · 2016

FractalNet: Ultra-Deep Neural Networks without Residuals

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 0 have no contract verdict; 4 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/1605.07648.json · badge: /api/ran/1605.07648.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
gustavla/fractalnet canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
jiye-ML/Classify_FRACTALNET reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
snf/keras-fractalnet reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 2 no recorded run
FunctionStatusWhere it lives
conv Ran
ran; our contract draft was wrong, not the code
gustavla/fractalnet/generation/fractalnet.py
code served (permissive licence) · get_code("cbc9a548b26d62c9")
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
from string import Template

COLORS = dict(black='0;30', darkgray='1;30', red='1;31', green='1;32', brown='0;33', yellow='1;33', blue='1;34', purple='1;35', cyan='1;36', white='1;37', reset='0')

DEFAULT_COLORIZE = False

def paint(s, color, colorize=DEFAULT_COLORIZE):
    if colorize:
        if color in COLORS:
            return '\x1b[{}m{}\x1b[0m'.format(COLORS[color], s)
        else:
            raise ValueError('Invalid color')
    else:
        return s

def conv(depth, red, num, dropout=0, bottom=None, separate=False):
    top = 'conv{}_{}'.format(depth, red)
    d = {}
    d['name'] = top
    if separate:
        top = 's_' + top
    d['top'] = top
    d['bottom'] = bottom
    d['num'] = num
    d['size'] = 3
    d['dropout'] = dropout
    d['pad'] = 1
    d['func'] = paint('conv', 'green')
    s = Template(':call $func $name $dropout $size $pad $num $bottom $top').substitute(d)
    return (s, top)

ARGS = {
    'depth': 1,
    'red': 2,
    'num': 64,
    'dropout': 0.5,
    'bottom': 'data',
    'separate': False
}
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 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
paint Ran
ran; our contract draft was wrong, not the code
gustavla/fractalnet/generation/fractalnet.py
code served (permissive licence) · get_code("407676128651cf4f")
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 sys

COLORS = dict(black='0;30', darkgray='1;30', red='1;31', green='1;32', brown='0;33', yellow='1;33', blue='1;34', purple='1;35', cyan='1;36', white='1;37', reset='0')

DEFAULT_COLORIZE = len(sys.argv) == 2 and sys.argv[1] == '--color'

def paint(s, color, colorize=DEFAULT_COLORIZE):
    if colorize:
        if color in COLORS:
            return '\x1b[{}m{}\x1b[0m'.format(COLORS[color], s)
        else:
            raise ValueError('Invalid color')
    else:
        return s

ARGS = {
    's': 'Hello World',
    'color': 'green',
    'colorize': False
}
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
pool Ran
ran; our contract draft was wrong, not the code
gustavla/fractalnet/generation/fractalnet.py
code served (permissive licence) · get_code("b3442826ef00d81b")
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 sys

COLORS = dict(black='0;30', darkgray='1;30', red='1;31', green='1;32', brown='0;33', yellow='1;33', blue='1;34', purple='1;35', cyan='1;36', white='1;37', reset='0')

DEFAULT_COLORIZE = len(sys.argv) == 2 and sys.argv[1] == '--color'

def paint(s, color, colorize=DEFAULT_COLORIZE):
    if colorize:
        if color in COLORS:
            return '\x1b[{}m{}\x1b[0m'.format(COLORS[color], s)
        else:
            raise ValueError('Invalid color')
    else:
        return s

def pool(depth, red, bottom=None, separate=False):
    top = 'pool{}_{}'.format(depth, red)
    d = {}
    if separate:
        top = 's_' + top
    d['top'] = top
    d['bottom'] = bottom
    d['func'] = paint('pool2', 'purple')
    s = Template(':call $func $bottom $top').substitute(d)
    return (s, top)

from string import Template

ARGS = {
    'depth': 2,
    'red': 3,
    'bottom': 'input_layer',
    'separate': False
}
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 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
tensor_shape Ran
ran; our contract draft was wrong, not the code
jiye-ML/Classify_FRACTALNET/src/fractal_block.py
pointer only (licence: NONE) · get_code("ff58e87bb50373d8")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, jiye-ML/Classify_FRACTALNET 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.

tensorflow_categorical Not yet run
unverified (harvested, no recorded run)
snf/keras-fractalnet/src/fractalnet.py
code served (permissive licence) · get_code("4de38f4b392c1c66")
How it ran

no invocation recorded.

theano_multinomial Not yet run
unverified (harvested, no recorded run)
snf/keras-fractalnet/src/fractalnet.py
code served (permissive licence) · get_code("c1545d890d27b807")
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, 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

We introduce a design strategy for neural network macro-architecture based on self-similarity. Repeated application of a simple expansion rule generates deep networks whose structural layouts are precisely truncated fractals. These networks contain interacting subpaths of different lengths, but do not include any pass-through or residual connections; every internal signal is transformed by a filter and nonlinearity before being seen by subsequent layers. In experiments, fractal networks match the excellent performance of standard residual networks on both CIFAR and ImageNet classification tasks, thereby demonstrating that residual representations may not be fundamental to the success of extremely deep convolutional neural networks. Rather, the key may be the ability to transition, during training, from effectively shallow to deep. We note similarities with student-teacher behavior and develop drop-path, a natural extension of dropout, to regularize co-adaptation of subpaths in fractal architectures. Such regularization allows extraction of high-performance fixed-depth subnetworks. Additionally, fractal networks exhibit an anytime property: shallow subnetworks provide a quick answer, while deeper subnetworks, with higher latency, provide a more accurate answer.

For agents

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

get_harvested_code_for_paper("1605.07648")
get_code_for_paper("1605.07648")
have("1605.07648")

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

curl https://syntology.ai/api/ran/1605.07648.json

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

[![Syntology run record](https://syntology.ai/api/ran/1605.07648.svg)](https://syntology.ai/paper/1605.07648)

Connect an agent — have() is free.