SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2301.13636 · 2023

Transport with Support: Data-Conditional Diffusion Bridges

arXiv · PDF · Open in the Atlas

Code that ran

6 of 9 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 6 have no contract verdict; 0 failures of our instrument; 3 no recorded run), measured 2026-11-21; 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/2301.13636.json · badge: /api/ran/2301.13636.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
aaltoml/iterative-smoothing-bridge canonical (classifier label)
classifier basis not recorded on the link
6 ran · 3 no recorded run
FunctionStatusWhere it lives
create_2d_obs_data Ran
ran on a synthesized input
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/load_2d.py
code served (permissive licence) · get_code("15a35513de3ab8cf")
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 numpy as np
import pandas as pd

torch.manual_seed(0)

# Create a minimal config object with the required nested structure
class FilterConfig:
    def __init__(self):
        self.n_steps = 5
        self.time_diff = 1.0
        self.time_forward = True
        self.n_ts = 2

class Config:
    def __init__(self):
        self.filter = FilterConfig()

# Create a minimal pandas dataframe that matches what the function expects
# The function expects:
# - data.iloc[:, :3] to work
# - data.times to exist
# - data['obs_0'] and data['obs_1'] to exist (n_dims = 3 - 1 = 2)
data_dict = {
    'times': [0.0, 1.0, 2.0, 3.0],
    'obs_0': ['1.5', '2.5', '3.5', '4.5'],
    'obs_1': ['5.5', '6.5', '7.5', '8.5']
}
data = pd.DataFrame(data_dict)

# Reshape data to have n_ts=2 time series
# Original shape is (4, 3), need to reshape to (2, -1, ...) via reshape(n_ts, -1, ...)
# So we need 4 rows total which gives us (2, 2) when reshaped
data = data.copy()

config = Config()

ARGS = {
    'data': data,
    'config': config,
    'device': 'cpu'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas, 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
generate_blocked_mnist Ran
ran on a synthesized input
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/torch_loaders.py
code served (permissive licence) · get_code("8747ad52775d85a6")
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 5, dtype family float, shape (2, 1, 1, 4, 6) 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)

# Analyze the function:
# - n_dim: used as *n_dim to unpack for torch.zeros, and n_dim[1] for indexing
#   This means n_dim should be a tuple/list of integers representing image dimensions
# - n_images: integer, number of images to generate
# - device: string, device specification (default 'cpu')

# The function creates tensors with shape (*n_dim, ...) so n_dim should be a tuple
# n_dim[1] is accessed, so it needs at least 2 elements
# Looking at the logic: block_imgs has shape (n_images, 1, *n_dim)
# and then indexing [:, :, :, :, :half_size] suggests spatial dimensions

# Let's use n_dim = (4, 6) to represent a 4x6 image (height x width)
n_dim = (4, 6)
n_images = 2
device = 'cpu'

ARGS = {
    'n_dim': n_dim,
    'n_images': n_images,
    'device': device
}
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 5, dtype family float, shape (2, 1, 1, 4, 6) — shape (2, 1, 1, 4, 6) 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_ssm_data Ran
ran on a synthesized input
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/ssm_data.py
code served (permissive licence) · get_code("394d25428260dc1f")
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 os
import pandas as pd
import torch
import numpy as np
import tempfile
import shutil

torch.manual_seed(0)

# Create a temporary directory structure with test data
temp_dir = tempfile.mkdtemp()
data_folder = os.path.join(temp_dir, 'data', 'test_model')
os.makedirs(data_folder, exist_ok=True)

# Create a small CSV file with observation data
# n_steps=100, so we need data that can be reshaped into (-1, 99, n_dims)
# Let's create 198 rows (2 sequences of 99 steps each) with 4 observation dimensions
n_sequences = 2
n_steps = 100
n_dims = 4
total_rows = n_sequences * (n_steps - 1)

data_dict = {'times': np.linspace(0, 1, total_rows)}
for i in range(n_dims):
    data_dict[f'obs_{i}'] = np.random.randn(total_rows)

df = pd.DataFrame(data_dict)
df.to_csv(os.path.join(data_folder, 'obs.csv'), index=False)

# Set up the arguments
ARGS = {
    'folder': temp_dir,
    'model_type': 'test_model',
    'device': torch.device('cpu'),
    'n_steps': 100
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas, 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
get_ssm_moments Ran
ran on a synthesized input
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/ssm_data.py
code served (permissive licence) · get_code("cea83c2691b500b3")
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 os
import pandas as pd
import torch
import numpy as np
import tempfile
import shutil

torch.manual_seed(0)

# Create a temporary directory structure
temp_dir = tempfile.mkdtemp()
data_folder = os.path.join(temp_dir, 'data', 'test_model_ssm')
os.makedirs(data_folder, exist_ok=True)

# Create a minimal CSV file that matches what the function expects
csv_data = {
    'times': [0.0, 1.0],
    'means': ['[1.0 2.0]', '[3.0 4.0]'],
    'vars': [
        '[[1.0 0.5]\n [0.5 2.0]]',
        '[[1.5 0.3]\n [0.3 2.5]]'
    ]
}
df = pd.DataFrame(csv_data)
csv_path = os.path.join(data_folder, 'moments_test_task.csv')
df.to_csv(csv_path, index=False)

ARGS = {
    'folder': temp_dir,
    'model_type': 'test_model',
    'device': 'cpu',
    'task': 'test_task'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas, 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
load_toy_data Ran
ran on a synthesized input
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/load_2d.py
code served (permissive licence) · get_code("53e63ed47d653e54")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from torch.utils.data import TensorDataset
from sklearn import datasets

torch.manual_seed(0)
np.random.seed(0)

ARGS = {
    'data_names': ['moons', 'circles'],
    'batch_size': 2,
    'device': 'cpu',
    'offset': 0.5,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
prep_ssm_data Ran
ran on a synthesized input
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/ssm_data.py
code served (permissive licence) · get_code("064b9557f441925c")
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

torch.manual_seed(0)

class FilterConfig:
    def __init__(self):
        self.n_steps = 10
        self.n_ts = 8
        self.n_obs = 2
        self.n_subs = 2
        self.time_forward = False

class Config:
    def __init__(self):
        self.filter = FilterConfig()

config = Config()

all_times = torch.randn(1, 10)
all_samples = torch.randn(8, 10)

ARGS = {
    "config": config,
    "all_times": all_times,
    "all_samples": all_samples
}
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
get_mnist_loader Not yet run
unverified (harvested, no recorded run)
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/torch_loaders.py
code served (permissive licence) · get_code("e736539cfed89058")
How it ran

no invocation recorded.

get_rna_data Not yet run
unverified (harvested, no recorded run)
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/torch_loaders.py
code served (permissive licence) · get_code("819a15285f541e22")
How it ran

no invocation recorded.

read_bridge_observations Not yet run
unverified (harvested, no recorded run)
aaltoml/iterative-smoothing-bridge/src/isb/data_loaders/load_2d.py
code served (permissive licence) · get_code("af261721dadd4c17")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 9 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-21.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The dynamic Schrödinger bridge problem provides an appealing setting for solving constrained time-series data generation tasks posed as optimal transport problems. It consists of learning non-linear diffusion processes using efficient iterative solvers. Recent works have demonstrated state-of-the-art results (eg. in modelling single-cell embryo RNA sequences or sampling from complex posteriors) but are limited to learning bridges with only initial and terminal constraints. Our work extends this paradigm by proposing the Iterative Smoothing Bridge (ISB). We integrate Bayesian filtering and optimal control into learning the diffusion process, enabling the generation of constrained stochastic processes governed by sparse observations at intermediate stages and terminal constraints. We assess the effectiveness of our method on synthetic and real-world data generation tasks and we show that the ISB generalises well to high-dimensional data, is computationally efficient, and provides accurate estimates of the marginals at intermediate and terminal times.

For agents

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

get_harvested_code_for_paper("2301.13636")
get_code_for_paper("2301.13636")
have("2301.13636")

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

curl https://syntology.ai/api/ran/2301.13636.json

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

[![Syntology run record](https://syntology.ai/api/ran/2301.13636.svg)](https://syntology.ai/paper/2301.13636)

Connect an agent — have() is free.