SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2110.11291 · 2021

Likelihood Training of Schrödinger Bridge using Forward-Backward SDEs Theory

arXiv · PDF · Open in the Atlas

Code that ran

9 of 11 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; 2 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 5 failures of our instrument; 2 no recorded run); 1 of them was also run on the host, not in the jail, by the peer-agreement lane that set its level, 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/2110.11291.json · badge: /api/ran/2110.11291.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
ghliu/sb-fbsde canonical (classifier label)
classifier basis not recorded on the link
9 ran · 2 no recorded run
FunctionStatusWhere it lives
sample_rademacher_like Ran
ran; our contract draft was wrong, not the code
ghliu/sb-fbsde/loss.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from BorealisAI/continuous-time-flow-process, is pointer only (licence: NOASSERTION) · get_code("7b0725913f7abbbb")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from BorealisAI/continuous-time-flow-process, 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.

blue Ran
ran; our contract draft was wrong, not the code
ghliu/sb-fbsde/util.py
code served (permissive licence) · get_code("831ff3e5265f6557")
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 termcolor

# The function takes 'content' as a parameter and converts it to string,
# then passes it to termcolor.colored with color 'blue' and bold attribute.
# We need to provide a value for 'content' that can be converted to string.

ARGS = {
    'content': 'Hello, World!'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
termcolor — 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
checkpoint Ran
ran, honoured the contract we drafted
ghliu/sb-fbsde/models/utils.py
code served (permissive licence) · get_code("aafea5a6cd6d98ab")
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, 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)

# Define a simple function that takes input tensors and returns output tensors
def simple_func(x, y):
    """A function that takes two tensors and returns their sum."""
    return x + y

# Create small input tensors
input_x = torch.randn(2, 4, requires_grad=True)
input_y = torch.randn(2, 4, requires_grad=True)

# Create parameter tensors (tensors the function depends on but doesn't take as explicit arguments)
param_1 = torch.randn(4, 3, requires_grad=True)
param_2 = torch.randn(3, 1, requires_grad=True)

# Define arguments for the checkpoint function
ARGS = {
    'func': simple_func,
    'inputs': [input_x, input_y],
    'params': [param_1, param_2],
    'flag': True
}
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, 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
get_data_dim Ran
ran on a synthesized input
ghliu/sb-fbsde/data.py
code served (permissive licence) · get_code("0cea5977ff26ee79")
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 numpy as np

torch.manual_seed(0)

ARGS = {
    'problem_name': 'mnist'
}
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 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
green Ran
ran; our contract draft was wrong, not the code
ghliu/sb-fbsde/util.py
code served (permissive licence) · get_code("6c22ac743d271a9e")
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 termcolor

def green(content):
    return termcolor.colored(str(content), 'green', attrs=['bold'])

# The function takes a single parameter 'content' and converts it to string,
# then colors it green with bold attribute using termcolor.colored()
# We can pass any object that can be converted to string

ARGS = {
    'content': 'Hello World'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
termcolor — 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
red Ran
ran; our contract draft was wrong, not the code
ghliu/sb-fbsde/util.py
code served (permissive licence) · get_code("f975c78dd7a0439b")
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 termcolor

# The function red() takes a 'content' parameter and converts it to a string,
# then colors it red with bold styling using termcolor.colored()

# We need to provide a value for 'content' that can be converted to string
# Let's use a simple string value

ARGS = {
    'content': 'Hello, World!'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
termcolor — 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
sample_e Ran
ran on a synthesized input
ghliu/sb-fbsde/loss.py
code served (permissive licence) · get_code("0601e2625342e738")
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, 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

torch.manual_seed(0)

class OptConfig:
    def __init__(self, noise_type):
        self.noise_type = noise_type

opt = OptConfig(noise_type='gaussian')
x = torch.randn(2, 4)

ARGS = {
    'opt': opt,
    'x': x
}
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, 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
sample_gaussian_like Ran
ran, honoured the contract we drafted
ghliu/sb-fbsde/loss.py
code served (permissive licence) · get_code("97f801e33d907e8a")
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, 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

torch.manual_seed(0)

# The function sample_gaussian_like takes a tensor y and returns
# torch.randn_like(y), which generates a tensor of random normal values
# with the same shape as y.
# We need to provide a tensor as the argument.

y = torch.randn(2, 4)

ARGS = {
    "y": y
}
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, 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
zero_module Ran
ran; our contract draft was wrong, not the code
ghliu/sb-fbsde/models/utils.py
code served (permissive licence) · get_code("129b804760b3115f")
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) and returned SimpleModule. 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 simple module with some parameters
class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
        self.conv = nn.Conv1d(2, 2, 3)
    
    def forward(self, x):
        return x

module = SimpleModule()

# Verify the module has non-zero parameters before zeroing
for p in module.parameters():
    assert p.abs().sum() > 0, "Parameters should be non-zero initially"

ARGS = {
    "module": module
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SimpleModule — a SimpleModule output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModule 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
avg_pool_nd Not yet run
unverified (harvested, no recorded run)
ghliu/sb-fbsde/models/utils.py
code served (permissive licence) · get_code("ecd0fc28815b65ae")
How it ran

no invocation recorded.

divergence_approx Not yet run
unverified (harvested, no recorded run)
ghliu/sb-fbsde/sde.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from BorealisAI/continuous-time-flow-process, is pointer only (licence: NOASSERTION) · get_code("f3c69a82cb618229")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from BorealisAI/continuous-time-flow-process, 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. 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, 9 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 0 the same month, 9 later) and 2 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
zero_module 196 other papers
3 dated earlier, 193 later · 49 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
2105.05233 (2021-05)
Diffusion Models Beat GANs on Image Synthesis
and 193 more not named here
avg_pool_nd 113 other papers
2 dated earlier, 111 later · 31 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
2111.11672 (2021-11)
Few-shot Image Generation with Mixup-based Distance Learning
and 110 more not named here
checkpoint 51 other papers
2 dated earlier, 49 later · 29 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2105.05233 (2021-05)
Diffusion Models Beat GANs on Image Synthesis
2209.15264 (2022-09)
Diffusion-based Image Translation using Disentangled Style and Content Representation
and 48 more not named here
sample_rademacher_like 6 other papers
4 dated earlier, 2 later · 3 only as an assembly brick
1810.01367 (2018-10)
FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models
1907.03907 (2019-07)
Latent ODEs for Irregularly-Sampled Time Series
2002.04461 (2020-02)
TrajectoryNet: A Dynamic Optimal Transport Network for Modeling Cellular Dynamics
and 3 more not named here
divergence_approx 5 other papers
4 dated earlier, 1 later · 3 only as an assembly brick
1810.01367 (2018-10)
FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models
1907.03907 (2019-07)
Latent ODEs for Irregularly-Sampled Time Series
2002.04461 (2020-02)
TrajectoryNet: A Dynamic Optimal Transport Network for Modeling Cellular Dynamics
and 2 more not named here
sample_gaussian_like 4 other papers
2 dated earlier, 2 later · 2 only as an assembly brick
1810.01367 (2018-10)
FFJORD: Free-form Continuous Dynamics for Scalable Reversible Generative Models
2002.04461 (2020-02)
TrajectoryNet: A Dynamic Optimal Transport Network for Modeling Cellular Dynamics
2305.07247 (2023-05)
Provably Convergent Schrödinger Bridge with Applications to Probabilistic Time Series Imputation
and 1 more not named here
blue 3 other papers
1 dated earlier, 2 later
1903.08642 (2019-03)
Photometric Mesh Optimization for Video-Aligned 3D Object Reconstruction
2303.01751 (2023-03)
Deep Momentum Multi-Marginal Schrödinger Bridge
2310.00108 (2023-10)
Practical Membership Inference Attacks Against Large-Scale Multi-Modal Models: A Pilot Study
green 3 other papers
1 dated earlier, 2 later
1903.08642 (2019-03)
Photometric Mesh Optimization for Video-Aligned 3D Object Reconstruction
2303.01751 (2023-03)
Deep Momentum Multi-Marginal Schrödinger Bridge
2310.00108 (2023-10)
Practical Membership Inference Attacks Against Large-Scale Multi-Modal Models: A Pilot Study
red 3 other papers
1 dated earlier, 2 later
1903.08642 (2019-03)
Photometric Mesh Optimization for Video-Aligned 3D Object Reconstruction
2303.01751 (2023-03)
Deep Momentum Multi-Marginal Schrödinger Bridge
2310.00108 (2023-10)
Practical Membership Inference Attacks Against Large-Scale Multi-Modal Models: A Pilot Study

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

Schrödinger Bridge (SB) is an entropy-regularized optimal transport problem that has received increasing attention in deep generative modeling for its mathematical flexibility compared to the Scored-based Generative Model (SGM). However, it remains unclear whether the optimization principle of SB relates to the modern training of deep generative models, which often rely on constructing log-likelihood objectives.This raises questions on the suitability of SB models as a principled alternative for generative applications. In this work, we present a novel computational framework for likelihood training of SB models grounded on Forward-Backward Stochastic Differential Equations Theory - a mathematical methodology appeared in stochastic optimal control that transforms the optimality condition of SB into a set of SDEs. Crucially, these SDEs can be used to construct the likelihood objectives for SB that, surprisingly, generalizes the ones for SGM as special cases. This leads to a new optimization principle that inherits the same SB optimality yet without losing applications of modern generative training techniques, and we show that the resulting training algorithm achieves comparable results on generating realistic images on MNIST, CelebA, and CIFAR10. Our code is available at https://github.com/ghliu/SB-FBSDE.

For agents

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

get_harvested_code_for_paper("2110.11291")
get_code_for_paper("2110.11291")
have("2110.11291")

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

curl https://syntology.ai/api/ran/2110.11291.json

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

[![Syntology run record](https://syntology.ai/api/ran/2110.11291.svg)](https://syntology.ai/paper/2110.11291)

Connect an agent — have() is free.