SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1905.10403 · 2019

Neural Jump Stochastic Differential Equations

arXiv · PDF · Open in the Atlas

Code that ran

4 of 5 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 2 have no contract verdict; 2 failures 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/1905.10403.json · badge: /api/ran/1905.10403.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
000Justin000/torchdiffeq canonical (classifier label)
classifier basis: census
4 ran · 1 no recorded run
FunctionStatusWhere it lives
compute_implicit_phi Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
000Justin000/torchdiffeq/torchdiffeq/_impl/adams.py
code served (permissive licence) · get_code("a269dfdac4e51213")
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 deque. 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 collections

torch.manual_seed(0)

# Analyze function usage:
# - explicit_phi: used with len() and indexing [j-1], elements are iterable (zip with implicit_phi[j-1])
# - f_n: appended to deque, used in zip operations, so should be iterable of numbers
# - k: used with min() and range()

# explicit_phi should be a list of tuples/iterables
# Each element represents phi values that can be subtracted from implicit_phi values
# Let's create a simple example with 3 explicit phi tuples, each with 4 elements

explicit_phi = [
    (1.0, 2.0, 3.0, 4.0),
    (0.5, 1.5, 2.5, 3.5),
    (0.2, 0.4, 0.6, 0.8),
]

# f_n should be an iterable of numbers matching the dimension of explicit_phi elements
f_n = (5.0, 6.0, 7.0, 8.0)

# k is the maximum length parameter
k = 5

ARGS = {
    "explicit_phi": explicit_phi,
    "f_n": f_n,
    "k": k,
}
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 deque — a deque output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a deque output the recorded ndim 2 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
g_and_explicit_phi Ran
ran; our contract draft was wrong, not the code
000Justin000/torchdiffeq/torchdiffeq/_impl/adams.py
code served (permissive licence) · get_code("94c5d03613990039")
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 collections

torch.manual_seed(0)

# Analyze the function to understand what's needed:
# - prev_t: indexed with [0] and [j] where j goes from 1 to k-1, so it's a list/tuple of tensors
# - next_t: a scalar tensor (used in arithmetic)
# - implicit_phi: indexed with [j] where j goes from 0 to k-1, each element is a tuple of tensors
# - k: an integer

# Build prev_t: list of k tensors
k = 3
prev_t = [torch.tensor(float(i), dtype=torch.float32) for i in range(k)]

# Build next_t: scalar tensor
next_t = torch.tensor(5.0, dtype=torch.float32)

# Build implicit_phi: list of k tuples, each tuple contains tensors
# The function does: iphi_ * beat_cast for iphi_ in implicit_phi[j]
# So each element should be a tuple of small tensors
implicit_phi = [
    (torch.randn(4, 3), torch.randn(4, 3)),
    (torch.randn(4, 3), torch.randn(4, 3)),
    (torch.randn(4, 3), torch.randn(4, 3)),
]

ARGS = {
    'prev_t': prev_t,
    'next_t': next_t,
    'implicit_phi': implicit_phi,
    'k': k,
}
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
rk4_alt_step_func Ran
ran on a synthesized input
000Justin000/torchdiffeq/torchdiffeq/_impl/rk_common.py
code served (permissive licence) · get_code("915f7ec0ef60ec9a")
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 3, 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

torch.manual_seed(0)

# The function expects:
# - func: a callable that takes (t, y) where y is a tuple and returns a tuple
# - t: a scalar time value
# - dt: a scalar time step
# - y: a tuple of tensors
# - k1: optional tuple of tensors (derivatives at current point)

# Create a simple ODE function that takes (t, y_tuple) and returns derivatives
def ode_func(t, y):
    """Simple ODE function: dy/dt = -y for each element in the tuple."""
    return tuple(-y_i for y_i in y)

# Create small initial state as a tuple of tensors
y_initial = (
    torch.randn(2, 4),  # batch of 2, dimension 4
    torch.randn(2, 4),  # second component
)

# Time parameters
t_current = 0.0
dt_step = 0.1

# Optionally pre-compute k1 to test the k1 parameter
k1_precomputed = ode_func(t_current, y_initial)

# Arguments for the function
ARGS = {
    "func": ode_func,
    "t": t_current,
    "dt": dt_step,
    "y": y_initial,
    "k1": k1_precomputed,
}
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, ndim 3, 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
rk4_step_func Ran
ran on a synthesized input
000Justin000/torchdiffeq/torchdiffeq/_impl/rk_common.py
code served (permissive licence) · get_code("3e2bf9b07fd9cfc5")
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

torch.manual_seed(0)

# Define a simple ODE function that takes (t, y) and returns derivatives
# y is expected to be a tuple of tensors
def ode_func(t, y):
    # y is a tuple of tensors, return tuple of same structure with derivatives
    return tuple(torch.randn_like(y_) * 0.1 for y_ in y)

# Create small test tensors for y (tuple of 2 tensors, each 2x4)
t = torch.tensor(0.0)
dt = torch.tensor(0.1)
y = (torch.randn(2, 4), torch.randn(2, 3))

# Optional: pre-compute k1 if needed
k1 = None

ARGS = {
    "func": ode_func,
    "t": t,
    "dt": dt,
    "y": y,
    "k1": k1
}
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
odeint_adjoint Not yet run
unverified (harvested, no recorded run)
000Justin000/torchdiffeq/torchdiffeq/_impl/adjoint.py
code served (permissive licence) · get_code("344fee401e9251cf")
How it ran

no invocation recorded.

Also present in other papers' code

Of 5 samples lifted from this paper's linked repositories, 5 carry a function identity and 0 do not. Of the 5 distinct functions with an identity, 5 are also present in other papers whose code we hold (5 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
rk4_alt_step_func 2 other papers
1 dated earlier, 1 later
1806.07366 (2018-06)
Neural Ordinary Differential Equations
2003.00868 (2020-03)
Differentiable Molecular Simulations for Control and Learning
compute_implicit_phi 1 other paper
1 dated earlier
1806.07366 (2018-06)
Neural Ordinary Differential Equations
g_and_explicit_phi 1 other paper
1 dated earlier
1806.07366 (2018-06)
Neural Ordinary Differential Equations
odeint_adjoint 1 other paper
1 dated earlier
1806.07366 (2018-06)
Neural Ordinary Differential Equations
rk4_step_func 1 other paper
1 dated earlier
1806.07366 (2018-06)
Neural Ordinary Differential Equations

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

Many time series are effectively generated by a combination of deterministic continuous flows along with discrete jumps sparked by stochastic events. However, we usually do not have the equation of motion describing the flows, or how they are affected by jumps. To this end, we introduce Neural Jump Stochastic Differential Equations that provide a data-driven approach to learn continuous and discrete dynamic behavior, i.e., hybrid systems that both flow and jump. Our approach extends the framework of Neural Ordinary Differential Equations with a stochastic process term that models discrete events. We then model temporal point processes with a piecewise-continuous latent trajectory, where the discontinuities are caused by stochastic events whose conditional intensity depends on the latent state. We demonstrate the predictive capabilities of our model on a range of synthetic and real-world marked point process datasets, including classical point processes (such as Hawkes processes), awards on Stack Overflow, medical records, and earthquake monitoring.

For agents

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

get_harvested_code_for_paper("1905.10403")
get_code_for_paper("1905.10403")
have("1905.10403")

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

curl https://syntology.ai/api/ran/1905.10403.json

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

[![Syntology run record](https://syntology.ai/api/ran/1905.10403.svg)](https://syntology.ai/paper/1905.10403)

Connect an agent — have() is free.