SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.14989 · ICLR · 2024

Stable Neural Stochastic Differential Equations in Analyzing Irregular Time Series Data

Yongkyung Oh, Dong-Young Lim, Sungil Kim

arXiv · PDF · Open in the Atlas

Code that ran

2 of 3 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; 1 honoured the contract we drafted, 0 violated it and 0 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/2402.14989.json · badge: /api/ran/2402.14989.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
yongkyung-oh/Stable-Neural-SDEs canonical (classifier label)
classifier basis: census
2 ran · 1 no recorded run
FunctionStatusWhere it lives
cdeint Ran
ran; our contract draft was wrong, not the code
yongkyung-oh/Stable-Neural-SDEs/benchmark_classification/controldiffeq/cdeint_module.py
code served (permissive licence) · get_code("a4dafc84cbeb2bb5")
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 (5, 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 func module that takes z and returns vector field
class SimpleFunc(nn.Module):
    def __init__(self, hidden_channels, input_channels):
        super(SimpleFunc, self).__init__()
        self.hidden_channels = hidden_channels
        self.input_channels = input_channels
        # Simple linear layer to map hidden -> (hidden x input)
        self.linear = nn.Linear(hidden_channels, hidden_channels * input_channels)
    
    def forward(self, z):
        # z shape: (..., hidden_channels)
        # Should return (..., hidden_channels, input_channels)
        batch_shape = z.shape[:-1]
        out = self.linear(z)
        # Reshape to (..., hidden_channels, input_channels)
        return out.reshape(*batch_shape, self.hidden_channels, self.input_channels)

# Define dX_dt - a callable that returns control gradients
class ControlGradient(nn.Module):
    def __init__(self, input_channels, batch_shape):
        super(ControlGradient, self).__init__()
        self.input_channels = input_channels
        self.batch_shape = batch_shape
    
    def forward(self, t):
        # t is a scalar tensor
        # Should return tensor of shape (..., input_channels)
        # where ... matches z0's batch dimensions
        return torch.ones(*self.batch_shape, self.input_channels)

# Set up dimensions
batch_size = 2
hidden_channels = 4
input_channels = 3

# Create initial state z0
z0 = torch.randn(batch_size, hidden_channels)

# Create the func module
func = SimpleFunc(hidden_channels, input_channels)

# Create the control gradient callable
batch_shape = (batch_size,)
dX_dt = ControlGradient(input_channels, batch_shape)

# Create time points to integrate over
t = torch.linspace(0, 1, 5)

# Build ARGS dictionary
ARGS = {
    'dX_dt': dX_dt,
    'z0': z0,
    'func': func,
    't': t,
    'adjoint': False,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torchdiffeq — 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 (5, 2, 4) — shape (5, 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_min_max Ran
ran, honoured the contract we drafted
yongkyung-oh/Stable-Neural-SDEs/benchmark_interpolation/physionet.py
code served (permissive licence) · get_code("581484c900e0ef72")
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 2, 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

torch.manual_seed(0)

# Create sample records - each record is a tuple of (record_id, tt, vals, mask, labels)
# vals: (time_steps, n_features)
# mask: (time_steps, n_features) with 1 for non-missing, 0 for missing
# tt: time points
# record_id: identifier
# labels: target labels

record1 = (
    "record_1",  # record_id
    torch.tensor([0.0, 1.0, 2.0]),  # tt (time points)
    torch.tensor([[1.5, 2.3, 0.9, 1.1],
                  [2.1, 1.8, 1.2, 0.8],
                  [1.9, 2.5, 0.7, 1.3]]).float(),  # vals (3 timesteps, 4 features)
    torch.tensor([[1, 1, 0, 1],
                  [1, 0, 1, 1],
                  [1, 1, 1, 0]]).float(),  # mask (3 timesteps, 4 features)
    torch.tensor([0])  # labels
)

record2 = (
    "record_2",  # record_id
    torch.tensor([0.0, 1.0]),  # tt
    torch.tensor([[0.5, 3.2, 1.1, 2.5],
                  [1.2, 2.9, 0.6, 1.9]]).float(),  # vals (2 timesteps, 4 features)
    torch.tensor([[1, 1, 1, 1],
                  [0, 1, 1, 1]]).float(),  # mask (2 timesteps, 4 features)
    torch.tensor([1])  # labels
)

records = [record1, record2]
device = torch.device('cpu')

ARGS = {
    'records': records,
    '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 tuple, ndim 2, 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
GRU_ODE Not yet run
unverified (harvested, no recorded run)
yongkyung-oh/Stable-Neural-SDEs/benchmark_classification/models/vector_fields.py
code served (permissive licence) · get_code("95c5d0ccacb5efc4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 3 samples lifted from this paper's linked repositories, 3 carry a function identity and 0 do not. Of the 3 distinct functions with an identity, 3 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 1 later) and 0 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
get_data_min_max 3 other papers
3 later
2412.12886 (2024-12)
TimeCHEAT: A Channel Harmony Strategy for Irregularly Sampled Multivariate Time Series Analysis
aaai_34173 (2025 (year only))
DualDynamics: Synergizing Implicit and Explicit Methods for Robust Irregular Time Series Analysis
2601.16516 (2026-01)
Rethinking Large Language Models For Irregular Time Series Classification In Critical Care
cdeint 2 other papers
2 dated earlier
2005.08926 (2020-05)
Neural Controlled Differential Equations for Irregular Time Series
2303.10909 (2023-03)
Graph Neural Rough Differential Equations for Traffic Forecasting
GRU_ODE 1 other paper
1 dated earlier
2005.08926 (2020-05)
Neural Controlled Differential Equations for Irregular Time Series

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

Irregular sampling intervals and missing values in real-world time series data present challenges for conventional methods that assume consistent intervals and complete data. Neural Ordinary Differential Equations (Neural ODEs) offer an alternative approach, utilizing neural networks combined with ODE solvers to learn continuous latent representations through parameterized vector fields. Neural Stochastic Differential Equations (Neural SDEs) extend Neural ODEs by incorporating a diffusion term, although this addition is not trivial, particularly when addressing irregular intervals and missing values. Consequently, careful design of drift and diffusion functions is crucial for maintaining stability and enhancing performance, while incautious choices can result in adverse properties such as the absence of strong solutions, stochastic destabilization, or unstable Euler discretizations, significantly affecting Neural SDEs' performance. In this study, we propose three stable classes of Neural SDEs: Langevin-type SDE, Linear Noise SDE, and Geometric SDE. Then, we rigorously demonstrate their robustness in maintaining excellent performance under distribution shift, while effectively preventing overfitting. To assess the effectiveness of our approach, we conduct extensive experiments on four benchmark datasets for interpolation, forecasting, and classification tasks, and analyze the robustness of our methods with 30 public datasets under different missing rates. Our results demonstrate the efficacy of the proposed method in handling real-world irregular time series data.

For agents

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

get_harvested_code_for_paper("2402.14989")
get_code_for_paper("2402.14989")
have("2402.14989")

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

curl https://syntology.ai/api/ran/2402.14989.json

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

[![Syntology run record](https://syntology.ai/api/ran/2402.14989.svg)](https://syntology.ai/paper/2402.14989)

Connect an agent — have() is free.