SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1703.00810 · 2017

Opening the Black Box of Deep Neural Networks via Information

arXiv · PDF · Open in the Atlas

Code that ran

16 of 17 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; 3 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 13 failures of our instrument; 1 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/1703.00810.json · badge: /api/ran/1703.00810.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
etherandrius/information-networks reimplementation (classifier label)
classifier basis: classifier_2025
6 ran · 0 no recorded run
copy not recorded community
not in the paper's declared repositories
3 ran · 0 no recorded run
alomrani/IDNNs reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
makezur/information_bottleneck_pytorch reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
ravidziv/IDNNs reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
sepehr-rasouli/ITRL reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
StephanLorenzen/ExactIBAnalysisInQNNs reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
FunctionStatusWhere it lives
binarize Ran
ran; our contract draft was wrong, not the code
etherandrius/information-networks/information/NaftaliTishby.py
code served (permissive licence) · get_code("0625e50998d14878")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned ndarray, ndim 2, dtype family float, shape (3, 1) 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

# Create a 2D numpy array with distinct dimensions
data = np.array([[1.0, 2.0, 3.0, 4.0],
                 [5.0, 6.0, 7.0, 8.0],
                 [9.0, 10.0, 11.0, 12.0]], dtype=np.float32)

ARGS = {
    'data': data
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (3, 1) — shape (3, 1) 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
build_dist Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from gtegner/mine-pytorch
pointer only · get_code("2e757a88026d14d3")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

calc_information_for_layer Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ravidziv/IDNNs/idnns/information/information_process.py
pointer only (licence: NOASSERTION) · get_code("c64726ea73437f06")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, ravidziv/IDNNs (NOASSERTION). 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.

calc_velocity Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
alomrani/IDNNs/idnns/plots/plot_figures.py
pointer only (licence: NOASSERTION) · get_code("c7ac117fe021d1f1")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, alomrani/IDNNs (NOASSERTION). 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.

ema Ran
ran, honoured the contract we drafted
this paper's copy was not recorded; identical code first harvested from gtegner/mine-pytorch
pointer only · get_code("a1073d8f05df7b5a")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

ema_loss Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from gtegner/mine-pytorch
pointer only · get_code("33a893f5d09f90f7")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

entropy Ran
ran, honoured the contract we drafted
sepehr-rasouli/ITRL/mnist_mi.py
pointer only (licence: NONE) · get_code("5fcc1acb6f40d03b")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, sepehr-rasouli/ITRL 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.

entropy_of_probabilities Ran
ran, honoured the contract we drafted
etherandrius/information-networks/information/NaftaliTishby.py
code served (permissive licence) · get_code("0a8afec82ca4566b")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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

# Create a valid probability distribution (sums to 1, all values between 0 and 1)
probabilities = np.array([0.25, 0.25, 0.25, 0.25])

ARGS = {
    "probabilities": probabilities
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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
extract_array Ran
ran; our contract draft was wrong, not the code
alomrani/IDNNs/idnns/plots/plot_figures.py
pointer only (licence: NOASSERTION) · get_code("0e5211fe3cbe7d5c")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, alomrani/IDNNs (NOASSERTION). 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.

filename Ran
ran; our contract draft was wrong, not the code
etherandrius/information-networks/main_as_if_random.py
code served (permissive licence) · get_code("583d051bd6b1090d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 torch

class Args:
    def __init__(self):
        self.train_size = 0.75
        self.epochs = 10
        self.data_set = 'mnist'
        self.activation = 'relu'
        self.batch_size = 32
        self.shape = 128
        self.epoch_list = (5, 10, 15)

args = Args()

ARGS = {'args': args}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
torch — 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
get_probabilities Ran
ran; our contract draft was wrong, not the code
etherandrius/information-networks/information/NaftaliTishby.py
code served (permissive licence) · get_code("43a34df93b4b67a4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 simple 1D numpy array with repeated values
# This will allow the function to compute probabilities of unique elements
data = np.array([1, 2, 2, 3, 3, 3, 1, 2, 4, 4])

ARGS = {
    'data': data
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — 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
get_probabilities Ran
ran; our contract draft was wrong, not the code
etherandrius/information-networks/main_as_if_random.py
code served (permissive licence) · get_code("31621f47d5230d8a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 numpy as np

# Create a simple numpy array with some repeated values
data = np.array([1, 2, 2, 3, 3, 3, 4, 4, 4, 4])

ARGS = {
    "data": data
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — 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
get_probabilities_map Ran
ran; our contract draft was wrong, not the code
etherandrius/information-networks/main_as_if_random.py
code served (permissive licence) · get_code("ad4ae5354a08a3bc")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) 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 numpy as np

# The function expects map_data to be a dictionary where each value is an array-like
# that can be passed to np.unique
map_data = {
    'feature1': np.array([1, 2, 2, 3, 3, 3, 1]),
    'feature2': np.array(['a', 'b', 'a', 'a', 'b', 'c']),
    'feature3': np.array([0.5, 0.5, 1.0, 1.5, 0.5, 1.0])
}

ARGS = {'map_data': map_data}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy — 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
load_figures Ran
ran; our contract draft was wrong, not the code
alomrani/IDNNs/idnns/plots/plot_figures.py
pointer only (licence: NOASSERTION) · get_code("d6626b73baa1f114")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, alomrani/IDNNs (NOASSERTION). 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.

load_tishby_toy_dataset Ran
ran; our contract draft was wrong, not the code
makezur/information_bottleneck_pytorch/pytorch_network.py
pointer only (licence: NONE) · get_code("9a35694d84f5e3b6")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, makezur/information_bottleneck_pytorch 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.

train_network Ran
ran; our contract draft was wrong, not the code
makezur/information_bottleneck_pytorch/pytorch_network.py
pointer only (licence: NONE) · get_code("f2a85ea2b597b0a1")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, makezur/information_bottleneck_pytorch 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.

compute_MI Not yet run
unverified (harvested, no recorded run)
StephanLorenzen/ExactIBAnalysisInQNNs/IB/experiment.py
pointer only (licence: GPL-3.0) · get_code("d5ce0afe906f0fd1")
How it ran

The body of this sample is withheld: GPL-3.0, as GitHub detects it on this paper's repository, StephanLorenzen/ExactIBAnalysisInQNNs, is not in the permissive set Syntology serves inline. 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.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 17 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 2 later) and 14 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
ema 3 other papers
1 dated earlier, 2 later
1606.03657 (2016-06)
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
1801.04062 (2018-01)
MINE: Mutual Information Neural Estimation
2311.01489 (2023-11)
Invariant Causal Imitation Learning for Generalizable Policies
ema_loss 3 other papers
1 dated earlier, 2 later
1606.03657 (2016-06)
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
1801.04062 (2018-01)
MINE: Mutual Information Neural Estimation
2311.01489 (2023-11)
Invariant Causal Imitation Learning for Generalizable Policies
build_dist 1 other paper
1 dated earlier
1606.03657 (2016-06)
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

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 their great success, there is still no comprehensive theoretical understanding of learning with Deep Neural Networks (DNNs) or their inner organization. Previous work proposed to analyze DNNs in the \textit{Information Plane}; i.e., the plane of the Mutual Information values that each layer preserves on the input and output variables. They suggested that the goal of the network is to optimize the Information Bottleneck (IB) tradeoff between compression and prediction, successively, for each layer. In this work we follow up on this idea and demonstrate the effectiveness of the Information-Plane visualization of DNNs. Our main results are: (i) most of the training epochs in standard DL are spent on {\emph compression} of the input to efficient representation and not on fitting the training labels. (ii) The representation compression phase begins when the training errors becomes small and the Stochastic Gradient Decent (SGD) epochs change from a fast drift to smaller training error into a stochastic relaxation, or random diffusion, constrained by the training error value. (iii) The converged layers lie on or very close to the Information Bottleneck (IB) theoretical bound, and the maps from the input to any hidden layer and from this hidden layer to the output satisfy the IB self-consistent equations. This generalization through noise mechanism is unique to Deep Neural Networks and absent in one layer networks. (iv) The training time is dramatically reduced when adding more hidden layers. Thus the main advantage of the hidden layers is computational. This can be explained by the reduced relaxation time, as this it scales super-linearly (exponentially for simple diffusion) with the information compression from the previous layer.

For agents

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

get_harvested_code_for_paper("1703.00810")
get_code_for_paper("1703.00810")
have("1703.00810")

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

curl https://syntology.ai/api/ran/1703.00810.json

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

[![Syntology run record](https://syntology.ai/api/ran/1703.00810.svg)](https://syntology.ai/paper/1703.00810)

Connect an agent — have() is free.