SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1909.03935 · 2019

GAN-Leaks: A Taxonomy of Membership Inference Attacks against Generative Models

arXiv · PDF · Open in the Atlas

Code that ran

13 of 22 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 7 have no contract verdict; 3 failures of our instrument; 9 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/1909.03935.json · badge: /api/ran/1909.03935.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
DingfanChen/GAN-Leaks canonical (classifier label)
classifier basis: census
13 ran · 9 no recorded run
FunctionStatusWhere it lives
adjust_dynamic_range Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
DingfanChen/GAN-Leaks/gan_models/pggan/misc.py
pointer only (licence: MIT); the stored copy get_code serves, from NVlabs/selfsupervised-denoising, is pointer only (licence: NOASSERTION) · get_code("90a97f8af1b8f269")
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 NVlabs/selfsupervised-denoising, and GitHub could not classify the licence of that repository (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.

conv_out_size_same Ran
ran, honoured the contract we drafted
DingfanChen/GAN-Leaks/gan_models/dcgan/model.py
pointer only (licence: MIT); the stored copy get_code serves, from xtarx/Unsupervised-Anomaly-Detection-with-Generative-Adversarial-Networks, is pointer only (licence: NONE) · get_code("40125e1efb67fa92")
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 xtarx/Unsupervised-Anomaly-Detection-with-Generative-Adversarial-Networks, 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.

create_image_grid Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/pggan/misc.py
code served (permissive licence) · get_code("cf03a3ce8cc3695e")
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 ndarray, ndim 3, dtype family float, shape (3, 4, 10) 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
import torch

torch.manual_seed(0)

# Create test images with shape (num_images, height, width) for grayscale
# or (num_images, channels, height, width) for color
# Using distinct dimensions: 2 images, 3 channels, 4 height, 5 width
images_4d = np.random.rand(2, 3, 4, 5).astype(np.float32)

ARGS = {
    "images": images_4d,
    "grid_size": None
}
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 ndarray, ndim 3, dtype family float, shape (3, 4, 10) — shape (3, 4, 10) 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
cset Ran
ran; our contract draft was wrong, not the code
DingfanChen/GAN-Leaks/gan_models/pggan/networks.py
pointer only (licence: MIT); the stored copy get_code serves, from LignumResearch/stylewood-model-usage, is pointer only (licence: NOASSERTION) · get_code("f7d82b02ff7c7fe1")
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 LignumResearch/stylewood-model-usage, and GitHub could not classify the licence of that repository (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.

expand_path Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/dcgan/utils.py
code served (permissive licence) · get_code("c2e6e9b41af6cd71")
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

ARGS = {
    "path": "~/test_directory/$HOME_VAR/file.txt"
}
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
flatten Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/pggan/tfutil.py
code served (permissive licence) · get_code("4e4ae70c11635981")
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 EagerTensor, ndim 1, dtype family float, shape (24,) 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 tensorflow as tf
import numpy as np

def flatten(x):
    with tf.name_scope('Flatten'):
        return tf.reshape(x, [-1])

# Create a small tensor with distinct dimensions
# Using shape (2, 4, 3) - batch_size=2, height=4, width=3
x = tf.constant(np.random.randn(2, 4, 3), dtype=tf.float32)

ARGS = {
    'x': x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 1, dtype family float, shape (24,) — shape (24,) 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
gen_random Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
DingfanChen/GAN-Leaks/gan_models/dcgan/model.py
code served (permissive licence) · get_code("040714a23128cf94")
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 ndarray, ndim 2, dtype family float, shape (2, 4) 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

ARGS = {
    'mode': 'normal01',
    'size': (2, 4)
}
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 ndarray, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
ignore_unknown_theano_network Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/pggan/legacy.py
code served (permissive licence) · get_code("2cd5be47220a5bc0")
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 torch
import torch.nn as nn

# The function takes a 'state' parameter which is a dictionary
# Looking at the function body:
# - It checks if 'version' is in state
# - If not, it accesses state['build_func_spec']['func']
# - It returns a modified state dict

# Create a state dict that triggers the print statement (no 'version' key)
state = {
    'build_func_spec': {
        'func': 'some_theano_function'
    }
}

ARGS = {
    'state': state
}
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 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
lerp Ran
ran, honoured the contract we drafted
DingfanChen/GAN-Leaks/gan_models/pggan/networks.py
pointer only (licence: MIT); the stored copy get_code serves, from thenhz/progressive_growing_of_gan, is pointer only (licence: NOASSERTION) · get_code("0c45399f3741921e")
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 thenhz/progressive_growing_of_gan, and GitHub could not classify the licence of that repository (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.

lerp_clip Ran
ran, honoured the contract we drafted
DingfanChen/GAN-Leaks/gan_models/pggan/networks.py
pointer only (licence: MIT); the stored copy get_code serves, from LignumResearch/stylewood-model-usage, is pointer only (licence: NOASSERTION) · get_code("6e81789f8f2068c8")
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 LignumResearch/stylewood-model-usage, and GitHub could not classify the licence of that repository (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.

read_image Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/dcgan/utils.py
code served (permissive licence) · get_code("2c432d530ff1bcb3")
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 ndarray, ndim 3, dtype family float, shape (64, 64, 3) 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
import PIL.Image
import tempfile
import os

# Create a temporary test image file
temp_dir = tempfile.mkdtemp()
test_image_path = os.path.join(temp_dir, "test_image.png")

# Create a 256x256 RGB image (larger than the 128x128 crop window)
test_image_array = np.random.randint(0, 256, (256, 256, 3), dtype=np.uint8)
test_image = PIL.Image.fromarray(test_image_array)
test_image.save(test_image_path)

ARGS = {
    "filepath": test_image_path,
    "resolution": 64,
    "cx": 89,
    "cy": 121
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family float, shape (64, 64, 3) — shape (64, 64, 3) 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
shape_to_list Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/pggan/tfutil.py
code served (permissive licence) · get_code("e963375233fe6dbe")
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 int; 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 torch.nn as nn

# The function accesses shape[i].value for each dimension in shape
# This suggests shape is a collection of objects with a .value attribute
# We need to create mock dimension objects

class MockDimension:
    def __init__(self, value):
        self.value = value

# Create a mock shape (like TensorFlow shape)
shape = [MockDimension(2), MockDimension(4), MockDimension(3)]

ARGS = {
    "shape": shape
}
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 int — 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
timestamp Ran
ran on a synthesized input
DingfanChen/GAN-Leaks/gan_models/dcgan/utils.py
code served (permissive licence) · get_code("221666d33a769cc9")
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 time
import datetime

ARGS = {
    's': '%Y%m%d.%H%M%S',
    'ts': time.time()
}
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
G_wgan_acgan Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/pggan/loss.py
code served (permissive licence) · get_code("82992ce1f4764e7b")
How it ran

no invocation recorded.

conv2d Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/dcgan/ops.py
code served (permissive licence) · get_code("5d3ea24b351a5a12")
How it ran

no invocation recorded.

conv_cond_concat Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/dcgan/ops.py
code served (permissive licence) · get_code("2906e84dbebc3e14")
How it ran

no invocation recorded.

create_celebahq Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/pggan/dataset_tool.py
code served (permissive licence) · get_code("c99617d8e7395b0f")
How it ran

no invocation recorded.

deconv2d Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/dcgan/ops.py
code served (permissive licence) · get_code("f1f0201f439514d5")
How it ran

no invocation recorded.

execute_cmdline Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/pggan/dataset_tool.py
code served (permissive licence) · get_code("81f0ed5010201398")
How it ran

no invocation recorded.

is_tf_expression Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/pggan/tfutil.py
code served (permissive licence) · get_code("574151db7f4e287b")
How it ran

no invocation recorded.

parse_tfrecord_np Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/pggan/dataset.py
code served (permissive licence) · get_code("9a489c1c146c4e18")
How it ran

no invocation recorded.

parse_tfrecord_tf Not yet run
unverified (harvested, no recorded run)
DingfanChen/GAN-Leaks/gan_models/pggan/dataset.py
code served (permissive licence) · get_code("490bcc28d486e9f7")
How it ran

no invocation recorded.

Also present in other papers' code

Of 22 samples lifted from this paper's linked repositories, 22 carry a function identity and 0 do not. Of the 22 distinct functions with an identity, 15 are also present in other papers whose code we hold (15 with at least one paper dated earlier, 0 the same month, 2 later) and 7 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
lerp 7 other papers
3 dated earlier, 4 later
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1802.08797 (2018-02)
Residual Dense Network for Image Super-Resolution
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
and 4 more not named here
conv_out_size_same 6 other papers
5 dated earlier, 1 later
1511.06434 (2015-11)
Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
1703.05921 (2017-03)
Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery
1804.04391 (2018-04)
MGGAN: Solving Mode Collapse using Manifold Guided Training
and 3 more not named here
cset 3 other papers
3 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1802.08797 (2018-02)
Residual Dense Network for Image Super-Resolution
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
lerp_clip 3 other papers
3 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1802.08797 (2018-02)
Residual Dense Network for Image Super-Resolution
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
G_wgan_acgan 2 other papers
2 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1802.08797 (2018-02)
Residual Dense Network for Image Super-Resolution
adjust_dynamic_range 2 other papers
2 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
1901.10277 (2019-01)
High-Quality Self-Supervised Deep Image Denoising
create_celebahq 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
create_image_grid 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
flatten 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
gen_random 1 other paper
1 dated earlier
1511.06434 (2015-11)
Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
ignore_unknown_theano_network 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
is_tf_expression 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
parse_tfrecord_np 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
parse_tfrecord_tf 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation
shape_to_list 1 other paper
1 dated earlier
1710.10196 (2017-10)
Progressive Growing of GANs for Improved Quality, Stability, and Variation

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

Deep learning has achieved overwhelming success, spanning from discriminative models to generative models. In particular, deep generative models have facilitated a new level of performance in a myriad of areas, ranging from media manipulation to sanitized dataset generation. Despite the great success, the potential risks of privacy breach caused by generative models have not been analyzed systematically. In this paper, we focus on membership inference attack against deep generative models that reveals information about the training data used for victim models. Specifically, we present the first taxonomy of membership inference attacks, encompassing not only existing attacks but also our novel ones. In addition, we propose the first generic attack model that can be instantiated in a large range of settings and is applicable to various kinds of deep generative models. Moreover, we provide a theoretically grounded attack calibration technique, which consistently boosts the attack performance in all cases, across different attack settings, data modalities, and training configurations. We complement the systematic analysis of attack performance by a comprehensive experimental study, that investigates the effectiveness of various attacks w.r.t. model type and training configurations, over three diverse application scenarios (i.e., images, medical data, and location data).

For agents

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

get_harvested_code_for_paper("1909.03935")
get_code_for_paper("1909.03935")
have("1909.03935")

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

curl https://syntology.ai/api/ran/1909.03935.json

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

[![Syntology run record](https://syntology.ai/api/ran/1909.03935.svg)](https://syntology.ai/paper/1909.03935)

Connect an agent — have() is free.