SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.14367 · CVPR · 2022

Thin-Plate Spline Motion Model for Image Animation

Jian Zhao, Hui Zhang

arXiv · PDF · Open in the Atlas

Code that ran

10 of 11 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (6 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 1 violated it and 7 have no contract verdict; 1 failure 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/2203.14367.json · badge: /api/ran/2203.14367.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
yoyo-nb/thin-plate-spline-motion-model marked official in Papers with Code
Papers with Code archive
10 ran · 1 no recorded run
FunctionStatusWhere it lives
AntiAliasInterpolation2d Ran
ran on a synthesized input
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("cc822200aa21f534")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 4, 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

# INIT_ARGS: parameters for __init__(channels, scale)
INIT_ARGS = {
    'channels': 3,
    'scale': 0.5
}

# ARGS: parameters for forward(input)
# Create a small input tensor with shape (batch, channels, height, width)
# channels must match INIT_ARGS['channels']
ARGS = {
    'input': torch.randn(2, 3, 8, 8)
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 4, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
Decoder Ran
ran on a synthesized input · constructed an object; no result was computed
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("b9c2907721f30e2d")
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 an instance of its own class, Decoder: the run constructed the object and did not call it, so no result was computed. 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)

# For Decoder, we need to understand what it expects:
# - block_expansion: int, scaling factor for channel dimensions
# - in_features: int, number of input features
# - num_blocks: int, number of decoder blocks (default 3)
# - max_features: int, maximum number of features (default 256)

# The forward method expects:
# - x: a list of tensors (it pops from this list)
# - mode: int (0 or 1)

# Let's trace through the logic:
# For num_blocks=3:
# - i goes 2, 1, 0 (reversed range)
# - At i=2: in_filters = 1 * min(256, block_expansion * 2^3) = block_expansion * 8
#           out_filters = min(256, block_expansion * 2^2) = block_expansion * 4
# - At i=1: in_filters = 2 * min(256, block_expansion * 2^2) = block_expansion * 8
#           out_filters = min(256, block_expansion * 2^1) = block_expansion * 2
# - At i=0: in_filters = 2 * min(256, block_expansion * 2^1) = block_expansion * 4
#           out_filters = min(256, block_expansion * 2^0) = block_expansion

# We need to construct input x as a list of tensors that will be popped.
# The forward method:
# 1. pops the first tensor (deepest feature map)
# 2. For each up_block, upsamples, concatenates with popped skip connection
# 3. This continues num_blocks times

# So we need num_blocks + 1 tensors in the list (one initial + num_blocks skips)

block_expansion = 4
in_features = 3
num_blocks = 3
max_features = 256

# Calculate expected channel sizes
# Initial bottleneck: 1 * min(256, 4 * 2^3) = 1 * min(256, 32) = 32
# After up_block 0 concat with skip (which is 4*4=16): 32 + 16 = 48
# After up_block 1 concat with skip (which is 4*2=8): 16 + 8 = 24
# After up_block 2 concat with skip (which is 4*1=4): 8 + 4 = 12

# Build input list in reverse order (will be popped)
# We need tensors such that after each up_block and concatenation we get right dims

# Initial tensor (bottleneck)
x_list = []

# Working backwards: at the last (first to be popped) we need the bottleneck
bottleneck_channels = 1 * min(max_features, block_expansion * 2 ** (num_blocks))
x_list.append(torch.randn(2, bottleneck_channels, 4, 4))

# Then we need skip connections in reverse order
for i in range(num_blocks - 1, -1, -1):
    skip_channels = min(max_features, block_expansion * 2 ** i)
    x_list.append(torch.randn(2, skip_channels, 8, 8))

ARGS = {
    'block_expansion': block_expansion,
    'in_features': in_features,
    'num_blocks': num_blocks,
    'max_features': max_features,
}
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 Decoder — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Decoder 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
DownBlock2d Ran
ran on a synthesized input · constructed an object; no result was computed
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("26141a5132dd41c8")
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 an instance of its own class, DownBlock2d: the run constructed the object and did not call it, so no result was computed. 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)

# DownBlock2d takes in_features, out_features, kernel_size, padding, and groups
# The forward method expects input tensor x of shape (batch, in_features, height, width)
# It applies Conv2d, InstanceNorm2d, ReLU, and AvgPool2d

# Create a small input tensor for testing: batch=2, in_features=4, height=8, width=8
input_tensor = torch.randn(2, 4, 8, 8)

ARGS = {
    'in_features': 4,
    'out_features': 3,
    'kernel_size': 3,
    'padding': 1,
    'groups': 1
}
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 DownBlock2d — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a DownBlock2d 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
Encoder Ran
ran on a synthesized input · constructed an object; no result was computed
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("7734c6dbfc47b9ec")
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 an instance of its own class, Encoder: the run constructed the object and did not call it, so no result was computed. 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)

ARGS = {
    "block_expansion": 64,
    "in_features": 3,
    "num_blocks": 3,
    "max_features": 256,
}
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 Encoder — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Encoder 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
Hourglass Ran
ran on a synthesized input · constructed an object; no result was computed
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("f13c1877591873b5")
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 an instance of its own class, Hourglass: the run constructed the object and did not call it, so no result was computed. 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)

ARGS = {
    'block_expansion': 64,
    'in_features': 3,
    'num_blocks': 3,
    'max_features': 256
}
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 Hourglass — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Hourglass 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
TPS Ran
ran on a synthesized input · constructed an object; no result was computed
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("f5c9c7ad1763b8ce")
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 an instance of its own class, TPS: the run constructed the object and did not call it, so no result was computed. 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)

# Test for 'kp' mode
bs = 2
gs = 3  # grid size for keypoints
n_kps = 4  # number of keypoints

# Create keypoint tensors for 'kp' mode
# kp_1 shape: (bs, gs, n_kps, 2)
kp_1 = torch.randn(bs, gs, n_kps, 2)
# kp_2 shape: (bs, gs, n_kps, 2)
kp_2 = torch.randn(bs, gs, n_kps, 2)

ARGS = {
    'mode': 'kp',
    'bs': bs,
    'kp_1': kp_1,
    'kp_2': kp_2,
}
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 TPS — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a TPS 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
UpBlock2d Ran
ran on a synthesized input · constructed an object; no result was computed
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("502353a4a40bcf3c")
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 an instance of its own class, UpBlock2d: the run constructed the object and did not call it, so no result was computed. 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)

ARGS = {
    'in_features': 16,
    'out_features': 32,
    'kernel_size': 3,
    'padding': 1,
    'groups': 1
}
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 UpBlock2d — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a UpBlock2d 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
from_homogeneous Ran
ran, violated the contract we drafted
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("a6ec956a2cd586e1")
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 (2, 2, 2) 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 from_homogeneous takes homogeneous coordinates (last dimension has 3 elements)
# and converts them to 2D Cartesian coordinates by dividing x,y by z (the last component).
# coordinates[..., :2] extracts first 2 dimensions
# coordinates[..., 2:3] extracts the 3rd dimension (z component)
# The division normalizes by z to get Cartesian coordinates

# Create a small tensor with homogeneous coordinates
# Shape: (batch=2, dim=4) where last dimension represents homogeneous coords [x, y, z, w]
# We need at least 3 elements in the last dimension for the slicing to work
coordinates = torch.tensor([
    [[1.0, 2.0, 2.0, 1.0], [3.0, 4.0, 2.0, 1.0]],
    [[5.0, 6.0, 1.0, 1.0], [7.0, 8.0, 4.0, 1.0]]
], dtype=torch.float32)

ARGS = {
    "coordinates": coordinates
}
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 Tensor, ndim 3, dtype family float, shape (2, 2, 2) — shape (2, 2, 2) 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
kp2gaussian Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("cba98ca3e7ba2298")
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 4, dtype family float, shape (2, 3, 4, 5) 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)

# kp2gaussian expects:
# - kp: a tensor of keypoints with shape (..., 2) where last dim is (x, y) coordinates
# - spatial_size: tuple of (height, width) 
# - kp_variance: a scalar for the gaussian variance

# Create a small keypoint tensor with batch and num_keypoints dimensions
# Shape: (batch=2, num_keypoints=3, coordinates=2)
kp = torch.randn(2, 3, 2)

# Spatial size as (height, width)
spatial_size = (4, 5)

# Variance parameter for gaussian
kp_variance = 0.1

ARGS = {
    "kp": kp,
    "spatial_size": spatial_size,
    "kp_variance": kp_variance
}
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 4, dtype family float, shape (2, 3, 4, 5) — shape (2, 3, 4, 5) 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
to_homogeneous Ran
ran, honoured the contract we drafted
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("c299e0261d0f079a")
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 float, shape (2, 5) 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)

# The function takes coordinates and converts them to homogeneous coordinates
# by appending a column of ones. It needs a tensor with at least 2 dimensions.
coordinates = torch.randn(2, 4)

ARGS = {
    "coordinates": coordinates,
}
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 float, shape (2, 5) — shape (2, 5) 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
DenseMotionNetwork Not yet run
unverified (harvested, no recorded run)
yoyo-nb/thin-plate-spline-motion-model/modules/dense_motion.py
code served (permissive licence) · get_code("e2e28d9d3f0b9c4f")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 11 distinct functions with an identity, 5 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 5 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.

FunctionAlso present inEarliest-datedOthers named
Encoder
brick of this paper's assembly: helper
4 other papers
2 dated earlier, 2 later · 2 only as an assembly brick
2003.00196 (2020-03)
First Order Motion Model for Image Animation
2104.11280 (2021-04)
Motion Representations for Articulated Animation
2204.05018 (2022-04)
Structure-Aware Motion Transfer with Deformable Anchor Model
and 1 more not named here
to_homogeneous
brick of this paper's assembly: helper
2 other papers
2 later · 1 only as an assembly brick
2204.05018 (2022-04)
Structure-Aware Motion Transfer with Deformable Anchor Model
2401.09146 (2024-01)
Continuous Piecewise-Affine Based Motion Model for Image Animation
AntiAliasInterpolation2d
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2204.05018 (2022-04)
Structure-Aware Motion Transfer with Deformable Anchor Model
from_homogeneous
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2204.05018 (2022-04)
Structure-Aware Motion Transfer with Deformable Anchor Model
kp2gaussian
brick of this paper's assembly: helper
1 other paper
1 later
2401.09146 (2024-01)
Continuous Piecewise-Affine Based Motion Model for Image Animation

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

Image animation brings life to the static object in the source image according to the driving video. Recent works attempt to perform motion transfer on arbitrary objects through unsupervised methods without using a priori knowledge. However, it remains a significant challenge for current unsupervised methods when there is a large pose gap between the objects in the source and driving images. In this paper, a new end-to-end unsupervised motion transfer framework is proposed to overcome such issues. Firstly, we propose thin-plate spline motion estimation to produce a more flexible optical flow, which warps the feature maps of the source image to the feature domain of the driving image. Secondly, in order to restore the missing regions more realistically, we leverage multi-resolution occlusion masks to achieve more effective feature fusion. Finally, additional auxiliary loss functions are designed to ensure that there is a clear division of labor in the network modules, encouraging the network to generate high-quality images. Our method 1 can animate a variety of objects, including talking faces, human bodies, and pixel animations. Experiments demonstrate that our method performs better on most benchmarks than the state of the art with visible improvements in motion-related metrics.

For agents

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

get_harvested_code_for_paper("2203.14367")
get_code_for_paper("2203.14367")
have("2203.14367")

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

curl https://syntology.ai/api/ran/2203.14367.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.14367.svg)](https://syntology.ai/paper/2203.14367)

Connect an agent — have() is free.