SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2106.13898 · 2021

Closed-form Continuous-time Neural Models

arXiv · PDF · Open in the Atlas

Code that ran

7 of 10 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 3 have no contract verdict; 3 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/2106.13898.json · badge: /api/ran/2106.13898.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
raminmh/CfC canonical (classifier label)
classifier basis: census
7 ran · 3 no recorded run
FunctionStatusWhere it lives
cut_in_sequences Ran
ran; our contract draft was wrong, not the code
raminmh/CfC/traffic_with_cfc.py
code served (permissive licence) · get_code("cb00e141c552f8f1")
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 tuple, 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 numpy as np

# Create small numpy arrays for x and y
# x and y need to have shape[0] > seq_len so the loop can iterate
# Using distinct sizes: 10 samples, 4 features
x = np.random.randn(10, 4).astype(np.float32)
y = np.random.randn(10, 4).astype(np.float32)

# seq_len must be less than x.shape[0]
seq_len = 3

# inc is the increment step (default 1)
inc = 1

ARGS = {
    "x": x,
    "y": y,
    "seq_len": seq_len,
    "inc": inc,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 4, 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
get_data_min_max Ran
ran on a synthesized input
raminmh/CfC/duv_physionet.py
code served (permissive licence) · get_code("4db90a7700379d69")
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 matching the expected format
# Each record is: (record_id, tt, vals, mask, labels)
# vals: shape (time_steps, n_features)
# mask: shape (time_steps, n_features) with 1 for observed, 0 for missing
# tt: time values
# record_id: identifier
# labels: labels

record_1 = (
    "record_001",  # record_id
    torch.arange(0, 3, dtype=torch.float32),  # tt (3 time steps)
    torch.randn(3, 4),  # vals (3 time steps, 4 features)
    torch.ones(3, 4),  # mask (all observed)
    torch.tensor([0, 1])  # labels
)

record_2 = (
    "record_002",
    torch.arange(0, 3, dtype=torch.float32),
    torch.randn(3, 4),  # vals (3 time steps, 4 features)
    torch.ones(3, 4),  # mask (all observed)
    torch.tensor([1, 0])
)

records = [record_1, record_2]

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
get_logger Ran
ran; our contract draft was wrong, not the code
raminmh/CfC/duv_utils.py
code served (permissive licence) · get_code("6e6961df228cbc5a")
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 RootLogger. 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 logging
import tempfile
import os

# Create temporary files for testing
temp_dir = tempfile.mkdtemp()
logpath = os.path.join(temp_dir, "test.log")
filepath = "main_script.py"

# Create a temporary package file to read
package_file = os.path.join(temp_dir, "package_module.py")
with open(package_file, 'w') as f:
    f.write("# This is a test package file\ndef hello():\n    pass\n")

ARGS = {
    'logpath': logpath,
    'filepath': filepath,
    'package_files': [package_file],
    'displaying': True,
    'saving': True,
    'debug': False
}
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 RootLogger — a RootLogger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a RootLogger 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
get_person_id Ran
ran, honoured the contract we drafted
raminmh/CfC/duv_person_activity.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from easonLuo2001/Hi-Patch, is pointer only (licence: NONE) · get_code("2e9dede6592a9d91")
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 easonLuo2001/Hi-Patch, and GitHub detected no licence in that repository when it was checked. 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.

lecun_tanh Ran
ran on a synthesized input
raminmh/CfC/tf_cfc.py
code served (permissive licence) · get_code("0f328141df551a0b")
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 EagerTensor, 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 tensorflow as tf
import numpy as np

# Create a small input tensor for the function
# The function takes x and applies lecun_tanh transformation
# x should be a tensorflow tensor
x = tf.constant([[1.0, 2.0, 3.0, 4.0],
                 [0.5, 1.5, 2.5, 3.5]], dtype=tf.float32)

ARGS = {
    'x': x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, 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
load_pickle Ran
ran; our contract draft was wrong, not the code
raminmh/CfC/duv_utils.py
code served (permissive licence) · get_code("93361336a4f19653")
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 dict. 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 pickle
import tempfile
import os

# Create a temporary pickle file with some test data
test_data = {
    'key1': 'value1',
    'key2': [1, 2, 3],
    'key3': {'nested': 'dict'}
}

# Create a temporary file and write pickle data to it
temp_fd, temp_path = tempfile.mkstemp(suffix='.pkl')
os.close(temp_fd)

with open(temp_path, 'wb') as f:
    pickle.dump(test_data, f)

# Define ARGS with the filename parameter
ARGS = {
    'filename': temp_path
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
variable_time_collate_fn2 Ran
ran on a synthesized input
raminmh/CfC/duv_physionet.py
code served (permissive licence) · get_code("6d746ef9f9fd0c28")
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 numpy as np
import torch

torch.manual_seed(0)

# Create a batch of time series data
# Each item: (record_id, tt, vals, mask, labels)
batch = []

# First item in batch
record_id_1 = 1
tt_1 = torch.tensor([0.0, 1.0, 2.0])  # T=3 time points
vals_1 = torch.randn(3, 4)  # T=3, D=4 variables
mask_1 = torch.ones(3, 4)  # T=3, D=4
labels_1 = torch.tensor(0)  # binary label
batch.append((record_id_1, tt_1, vals_1, mask_1, labels_1))

# Second item in batch
record_id_2 = 2
tt_2 = torch.tensor([0.0, 0.5, 1.5, 2.5])  # T=4 time points
vals_2 = torch.randn(4, 4)  # T=4, D=4 variables
mask_2 = torch.ones(4, 4)  # T=4, D=4
labels_2 = torch.tensor(1)  # binary label
batch.append((record_id_2, tt_2, vals_2, mask_2, labels_2))

# Create data_min and data_max for normalization
data_min = torch.zeros(4)
data_max = torch.ones(4)

# Create args dict
args = {}

ARGS = {
    'batch': batch,
    'args': args,
    'device': torch.device('cpu'),
    'data_type': 'train',
    'data_min': data_min,
    'data_max': data_max
}
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 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_person_dataset Not yet run
unverified (harvested, no recorded run)
raminmh/CfC/duv_person_activity.py
code served (permissive licence) · get_code("4e8272b4a1d5ad43")
How it ran

no invocation recorded.

get_physio Not yet run
unverified (harvested, no recorded run)
raminmh/CfC/duv_physionet.py
code served (permissive licence) · get_code("b25e22b146a7b1f9")
How it ran

no invocation recorded.

make_dataset Not yet run
unverified (harvested, no recorded run)
raminmh/CfC/duv_utils.py
code served (permissive licence) · get_code("03cd51541800f514")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 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, 4 later) and 5 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_logger 2 other papers
2 later
2204.09633 (2022-04)
SurvLatent ODE : A Neural ODE based time-to-event model with competing risks for longitudinal data improves cancer-associated Venous Thromboembolism (VTE) prediction
2601.19040 (2026-01)
OATS: Online Data Augmentation for Time Series Foundation Models
get_person_id 2 other papers
2 later
2604.27967 (2026-04)
Differentiable latent structure discovery for interpretable forecasting in clinical time series
2606.09030 (2026-06)
TRIAGE: Dialectical Reasoning for Explainable Risk Prediction on Irregularly Sampled Medical Time Series with LLMs
load_pickle 2 other papers
2 later
2204.09633 (2022-04)
SurvLatent ODE : A Neural ODE based time-to-event model with competing risks for longitudinal data improves cancer-associated Venous Thromboembolism (VTE) prediction
2601.19040 (2026-01)
OATS: Online Data Augmentation for Time Series Foundation Models
make_dataset 2 other papers
2 later
2204.09633 (2022-04)
SurvLatent ODE : A Neural ODE based time-to-event model with competing risks for longitudinal data improves cancer-associated Venous Thromboembolism (VTE) prediction
2601.19040 (2026-01)
OATS: Online Data Augmentation for Time Series Foundation Models
cut_in_sequences 1 other paper
1 dated earlier
2006.04439 (2020-06)
Liquid Time-constant Networks

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

Continuous-time neural processes are performant sequential decision-makers that are built by differential equations (DE). However, their expressive power when they are deployed on computers is bottlenecked by numerical DE solvers. This limitation has significantly slowed down the scaling and understanding of numerous natural physical phenomena such as the dynamics of nervous systems. Ideally, we would circumvent this bottleneck by solving the given dynamical system in closed form. This is known to be intractable in general. Here, we show it is possible to closely approximate the interaction between neurons and synapses -- the building blocks of natural and artificial neural networks -- constructed by liquid time-constant networks (LTCs) efficiently in closed-form. To this end, we compute a tightly-bounded approximation of the solution of an integral appearing in LTCs' dynamics, that has had no known closed-form solution so far. This closed-form solution substantially impacts the design of continuous-time and continuous-depth neural models; for instance, since time appears explicitly in closed-form, the formulation relaxes the need for complex numerical solvers. Consequently, we obtain models that are between one and five orders of magnitude faster in training and inference compared to differential equation-based counterparts. More importantly, in contrast to ODE-based continuous networks, closed-form networks can scale remarkably well compared to other deep learning instances. Lastly, as these models are derived from liquid networks, they show remarkable performance in time series modeling, compared to advanced recurrent models.

For agents

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

get_harvested_code_for_paper("2106.13898")
get_code_for_paper("2106.13898")
have("2106.13898")

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

curl https://syntology.ai/api/ran/2106.13898.json

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

[![Syntology run record](https://syntology.ai/api/ran/2106.13898.svg)](https://syntology.ai/paper/2106.13898)

Connect an agent — have() is free.