SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1910.02551 · 2019

Soft-Label Dataset Distillation and Text Dataset Distillation

arXiv · PDF · Open in the Atlas

Code that ran

10 of 13 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 10 have no contract verdict; 0 failures of our instrument; 3 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/1910.02551.json · badge: /api/ran/1910.02551.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
ilia10000/dataset-distillation canonical (classifier label)
classifier basis: census
9 ran · 3 no recorded run
harshp1802/text-dataset-distillation community (classifier label)
classifier basis not recorded on the link
1 ran · 0 no recorded run
FunctionStatusWhere it lives
clone_tuple Ran
ran on a synthesized input
ilia10000/dataset-distillation/networks/utils.py
code served (permissive licence) · get_code("9ff25651ae35371b")
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. 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

torch.manual_seed(0)

# Create a tuple of tensors with different requires_grad settings
tensor1 = torch.randn(2, 4, requires_grad=True)
tensor2 = torch.randn(2, 4, requires_grad=False)
tensor3 = torch.randn(2, 4, requires_grad=True)

tensors = (tensor1, tensor2, tensor3)

ARGS = {
    "tensors": tensors,
    "requires_grad": None
}
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 — 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
cross_entropy Ran
ran on a synthesized input
ilia10000/dataset-distillation/basics.py
code served (permissive licence) · get_code("971a7e33288dbfc5")
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 Tensor, 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 torch.nn

torch.manual_seed(0)

# pred: input tensor to LogSoftmax, shape (batch_size, num_classes)
# The function applies LogSoftmax, multiplies by soft_targets, and reduces
pred = torch.randn(2, 4, requires_grad=True)

# soft_targets: soft probability targets, same shape as pred
# Used in element-wise multiplication: -soft_targets * logsoftmax(pred)
soft_targets = torch.softmax(torch.randn(2, 4), dim=1)

ARGS = {
    "pred": pred,
    "soft_targets": soft_targets,
}
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 Tensor, 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
encode Ran
ran on a synthesized input
ilia10000/dataset-distillation/utils/baselines.py
code served (permissive licence) · get_code("f5202ff9824f30c7")
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 Tensor, ndim 3, dtype family float, shape (4, 1, 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 torch
import torch.nn as nn

torch.manual_seed(0)

class State:
    def __init__(self):
        self.ntoken = 10
        self.ninp = 4
        self.device = 'cpu'
        self.pretrained_vec = torch.randn(10, 4)

state = State()

d = torch.tensor([0, 1, 2, 3], dtype=torch.long)

ARGS = {
    'd': d,
    '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 Tensor, ndim 3, dtype family float, shape (4, 1, 4) — shape (4, 1, 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
get_baseline_label_for_one_step Ran
ran on a synthesized input
ilia10000/dataset-distillation/utils/baselines.py
code served (permissive licence) · get_code("8f695f4fc33f3364")
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 Tensor, ndim 2, dtype family int, shape (6, 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 torch

torch.manual_seed(0)

class State:
    def __init__(self, num_classes, init_labels, distilled_images_per_class_per_step, device, mode, attack_class=None, target_class=None):
        self.num_classes = num_classes
        self.init_labels = init_labels
        self.distilled_images_per_class_per_step = distilled_images_per_class_per_step
        self.device = device
        self.mode = mode
        self.attack_class = attack_class
        self.target_class = target_class

state = State(
    num_classes=3,
    init_labels=[0, 1, 2],
    distilled_images_per_class_per_step=2,
    device='cpu',
    mode='normal',
    attack_class=0,
    target_class=torch.tensor([1, 0, 0])
)

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 Tensor, ndim 2, dtype family int, shape (6, 3) — shape (6, 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
init_weights Ran
ran on a synthesized input
ilia10000/dataset-distillation/networks/utils.py
code served (permissive licence) · get_code("cabbd274bc3fcb07")
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 SimpleNet. 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
from torch.nn import init

torch.manual_seed(0)

# Create a simple state object with init parameters
class State:
    def __init__(self, init_type, init_param, pretrained_vec=None, learnable_embedding=True):
        self.init = init_type
        self.init_param = init_param
        self.pretrained_vec = pretrained_vec
        self.learnable_embedding = learnable_embedding

# Create a simple network with Conv, Linear, RNN, Norm, and Embedding layers
class SimpleNet(nn.Module):
    def __init__(self):
        super(SimpleNet, self).__init__()
        self.conv1 = nn.Conv2d(3, 4, kernel_size=3)
        self.linear1 = nn.Linear(4, 8)
        self.rnn = nn.LSTM(input_size=4, hidden_size=8, num_layers=1)
        self.norm = nn.BatchNorm1d(8)
        self.embedding = nn.Embedding(num_embeddings=10, embedding_dim=4)
    
    def forward(self, x):
        return x

# Create pretrained embedding vector
pretrained_vec = torch.randn(10, 4)

# Create state with different init types to test various branches
state = State(init_type='normal', init_param=0.02, pretrained_vec=pretrained_vec, learnable_embedding=True)

# Create network
net = SimpleNet()

ARGS = {
    'net': net,
    'state': state
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SimpleNet — a SimpleNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleNet 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
init_weights Ran
ran on a synthesized input
harshp1802/text-dataset-distillation/networks/utils.py
code served (permissive licence) · get_code("e0e7604fc1f63651")
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 MinimalNet. 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
from torch.nn import init

torch.manual_seed(0)

# Create a simple state object with required attributes
class State:
    def __init__(self):
        self.init = 'xavier'
        self.init_param = 1.0
        self.pretrained_vec = torch.randn(10, 8)
        self.learnable_embedding = True

# Build a minimal network that exercises the init_weights function
class MinimalNet(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 4, kernel_size=3)
        self.linear1 = nn.Linear(4, 2)
        self.norm1 = nn.BatchNorm2d(4)
        self.embedding = nn.Embedding(10, 8)
        self.lstm = nn.LSTM(input_size=4, hidden_size=6, num_layers=1, batch_first=True)
        self.rnn = nn.RNN(input_size=4, hidden_size=6, num_layers=1, batch_first=True)
    
    def forward(self, x):
        return x

# Instantiate network and state
net = MinimalNet()
state = State()

ARGS = {
    'net': net,
    'state': state
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type MinimalNet — a MinimalNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a MinimalNet 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
read_bndbox Ran
ran on a synthesized input
ilia10000/dataset-distillation/datasets/pascal_voc.py
code served (permissive licence) · get_code("fdee81a9e734615a")
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. 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
import os.path
import tempfile
import xml.etree.ElementTree as ET

# Create a temporary directory structure mimicking VOC dataset
temp_root = tempfile.mkdtemp()
voc_dir = os.path.join(temp_root, 'VOCdevkit', 'VOC2012', 'Annotations')
os.makedirs(voc_dir, exist_ok=True)

# Create sample XML annotation files
def create_xml_file(filepath, image_name, objects):
    root = ET.Element('annotation')
    
    filename_elem = ET.SubElement(root, 'filename')
    filename_elem.text = image_name
    
    for obj_name, xmin, ymin, xmax, ymax in objects:
        obj_elem = ET.SubElement(root, 'object')
        
        name_elem = ET.SubElement(obj_elem, 'name')
        name_elem.text = obj_name
        
        bndbox_elem = ET.SubElement(obj_elem, 'bndbox')
        
        xmin_elem = ET.SubElement(bndbox_elem, 'xmin')
        xmin_elem.text = str(xmin)
        
        ymin_elem = ET.SubElement(bndbox_elem, 'ymin')
        ymin_elem.text = str(ymin)
        
        xmax_elem = ET.SubElement(bndbox_elem, 'xmax')
        xmax_elem.text = str(xmax)
        
        ymax_elem = ET.SubElement(bndbox_elem, 'ymax')
        ymax_elem.text = str(ymax)
    
    tree = ET.ElementTree(root)
    tree.write(filepath)

# Create first XML file
create_xml_file(
    os.path.join(voc_dir, '000001.xml'),
    '000001.jpg',
    [('cat', 10, 20, 100, 150), ('dog', 50, 60, 120, 180)]
)

# Create second XML file
create_xml_file(
    os.path.join(voc_dir, '000002.xml'),
    '000002.jpg',
    [('person', 5, 10, 200, 300)]
)

ARGS = {
    'root': temp_root,
    'dataset': 'VOC2012',
    'paths': ['000001', '000002']
}
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 list — a list 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
to_np Ran
ran on a synthesized input
ilia10000/dataset-distillation/utils/io.py
code served (permissive licence) · get_code("0250946399ff305f")
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. 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

torch.manual_seed(0)

# Create sample steps with torch tensors
# Each step is a tuple of (data, label, lr)
# data: 4D tensor (batch, channels, height, width)
# label: 1D or 2D tensor
# lr: scalar tensor or None

data1 = torch.randn(2, 3, 4, 5)  # batch=2, channels=3, height=4, width=5
label1 = torch.randint(0, 10, (2,))
lr1 = torch.tensor(0.001)

data2 = torch.randn(2, 3, 4, 5)
label2 = torch.randint(0, 10, (2,))
lr2 = None

steps = [
    (data1, label1, lr1),
    (data2, label2, lr2)
]

ARGS = {
    "steps": steps
}
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 list — a list 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
to_torch Ran
ran on a synthesized input
ilia10000/dataset-distillation/utils/io.py
code served (permissive licence) · get_code("63eebd9902e8150b")
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 3, 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 numpy as np

torch.manual_seed(0)

# The function expects np_steps to be a list of tuples
# Each tuple contains tensor-like objects that can be converted to torch tensors
# The function checks the first element of the first step to determine if conversion is needed

# Create sample steps: list of tuples containing numpy arrays and torch tensors
np_steps = [
    (np.array([1.0, 2.0, 3.0, 4.0]), np.array([5.0, 6.0, 7.0, 8.0])),
    (np.array([9.0, 10.0, 11.0, 12.0]), np.array([13.0, 14.0, 15.0, 16.0])),
]

device = torch.device('cpu')

ARGS = {
    'np_steps': np_steps,
    'device': device,
}
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 list, ndim 3, 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
xentropy_cost Ran
ran on a synthesized input
ilia10000/dataset-distillation/basics.py
code served (permissive licence) · get_code("f26548178c3935fa")
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 Tensor, 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

torch.manual_seed(0)

# x_target and x_pred must have the same size
# The function applies log to x_pred and multiplies element-wise with x_target
# Then sums all values
# These should be probability distributions or similar (positive values for log to work)

batch_size = 2
num_classes = 4

x_target = torch.tensor([
    [1.0, 0.0, 0.0, 0.0],
    [0.0, 1.0, 0.0, 0.0]
])

x_pred = torch.tensor([
    [0.7, 0.1, 0.1, 0.1],
    [0.1, 0.8, 0.05, 0.05]
])

ARGS = {
    'x_target': x_target,
    'x_pred': x_pred
}
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 Tensor, 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
load_results Not yet run
unverified (harvested, no recorded run)
ilia10000/dataset-distillation/utils/io.py
code served (permissive licence) · get_code("062128c654a29276")
How it ran

no invocation recorded.

random_train Not yet run
unverified (harvested, no recorded run)
ilia10000/dataset-distillation/utils/baselines.py
code served (permissive licence) · get_code("c02233f6d63439a8")
How it ran

no invocation recorded.

read_split Not yet run
unverified (harvested, no recorded run)
ilia10000/dataset-distillation/datasets/pascal_voc.py
code served (permissive licence) · get_code("4cd96988611c685e")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 distinct functions with an identity, 5 are also present in other papers whose code we hold (5 with at least one paper dated earlier, 0 the same month, 2 later) and 8 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
clone_tuple 2 other papers
1 dated earlier, 1 later
1811.10959 (2018-11)
Dataset Distillation
2301.01197 (2023-01)
Backdoor Attacks Against Dataset Distillation
read_bndbox 2 other papers
1 dated earlier, 1 later
1811.10959 (2018-11)
Dataset Distillation
2301.01197 (2023-01)
Backdoor Attacks Against Dataset Distillation
load_results 1 other paper
1 dated earlier
1811.10959 (2018-11)
Dataset Distillation
to_np 1 other paper
1 dated earlier
1811.10959 (2018-11)
Dataset Distillation
to_torch 1 other paper
1 dated earlier
1811.10959 (2018-11)
Dataset Distillation

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

Dataset distillation is a method for reducing dataset sizes by learning a small number of synthetic samples containing all the information of a large dataset. This has several benefits like speeding up model training, reducing energy consumption, and reducing required storage space. Currently, each synthetic sample is assigned a single `hard' label, and also, dataset distillation can currently only be used with image data. We propose to simultaneously distill both images and their labels, thus assigning each synthetic sample a `soft' label (a distribution of labels). Our algorithm increases accuracy by 2-4% over the original algorithm for several image classification tasks. Using `soft' labels also enables distilled datasets to consist of fewer samples than there are classes as each sample can encode information for multiple classes. For example, training a LeNet model with 10 distilled images (one per class) results in over 96% accuracy on MNIST, and almost 92% accuracy when trained on just 5 distilled images. We also extend the dataset distillation algorithm to distill sequential datasets including texts. We demonstrate that text distillation outperforms other methods across multiple datasets. For example, models attain almost their original accuracy on the IMDB sentiment analysis task using just 20 distilled sentences. Our code can be found at $\href{https://github.com/ilia10000/dataset-distillation}{\text{https://github.com/ilia10000/dataset-distillation}}$.

For agents

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

get_harvested_code_for_paper("1910.02551")
get_code_for_paper("1910.02551")
have("1910.02551")

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

curl https://syntology.ai/api/ran/1910.02551.json

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

[![Syntology run record](https://syntology.ai/api/ran/1910.02551.svg)](https://syntology.ai/paper/1910.02551)

Connect an agent — have() is free.