SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1905.08494 · 2019

Deep Signature Transforms

arXiv · PDF · Open in the Atlas

Code that ran

11 of 14 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; 3 honoured the contract we drafted, 0 violated it and 7 have no contract verdict; 1 failure 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/1905.08494.json · badge: /api/ran/1905.08494.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
patrick-kidger/Deep-Signatures canonical (classifier label)
classifier basis: census
4 ran · 2 no recorded run
patrick-kidger/Deep-Signature-Transforms canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
anh-tong/signax community (classifier label)
classifier basis not recorded on the link
5 ran · 1 no recorded run
FunctionStatusWhere it lives
addcmul Ran
ran on a synthesized input
anh-tong/signax/src/signax/tensor_ops.py
code served (permissive licence) · get_code("32d0a5a694aa8db3")
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 ArrayImpl, ndim 3, dtype family float, shape (4, 4, 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 jax
import jax.numpy as jnp

# Create small JAX arrays for testing
# x: size=(n,n,...,n), ndim=ndim_x - let's use ndim_x=2, n=4
x = jnp.ones((4, 4))

# y: size=(n,n,...,n), ndim=ndim_y - let's use ndim_y=2, n=4
y = jnp.ones((4, 4))

# z: size=(n,n,...,n), ndim=ndim_z - let's use ndim_z=1, n=4
z = jnp.ones((4,))

ARGS = {
    "x": x,
    "y": y,
    "z": z,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 3, dtype family float, shape (4, 4, 4) — shape (4, 4, 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
compress Ran
ran on a synthesized input
anh-tong/signax/src/signax/utils.py
code served (permissive licence) · get_code("240a22a9a6f347bd")
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 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 jax
import jax.numpy as jnp

# Set seed for reproducibility
jax.config.update('jax_platform_name', 'cpu')

# Create small test tensors
# input: list of m-level tensors, each of shape (dim, dim, ..., dim)
# For testing, we'll create two 3D tensors with dim=4

dim = 4

# First tensor: shape (4, 4, 4)
input_tensor_1 = jnp.arange(64, dtype=jnp.float32).reshape((4, 4, 4))

# Second tensor: shape (4, 4, 4)
input_tensor_2 = jnp.arange(64, 128, dtype=jnp.float32).reshape((4, 4, 4))

input_list = [input_tensor_1, input_tensor_2]

# indices: list of index arrays
# Each indices array has shape (dim, n) where n <= ndim
# For 3D tensors, ndim=3, so n can be at most 3
# Let's create indices of shape (4, 2) for selecting 2-level entries

indices_1 = jnp.array([
    [0, 1, 2, 3],
    [1, 2, 3, 0]
], dtype=jnp.int32).T  # Shape: (4, 2)

indices_2 = jnp.array([
    [3, 2, 1, 0],
    [0, 1, 2, 3]
], dtype=jnp.int32).T  # Shape: (4, 2)

indices_list = [indices_1, indices_2]

ARGS = {
    "input": input_list,
    "indices": indices_list,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family float — 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
convert_to_tensor Ran
ran on a synthesized input
patrick-kidger/Deep-Signatures/packages/candle/tracing.py
code served (permissive licence) · get_code("98d8af89d9aae571")
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)

# Test case 1: tensor input (should be returned as-is)
tensor_input = torch.randn(2, 4)

# Test case 2: list input (should be converted to tensor with kwargs)
list_input = [[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]]

# For the first call, we'll use the tensor input
ARGS = {
    "tensor": tensor_input,
}
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
count_parameters Ran
ran, honoured the contract we drafted
patrick-kidger/Deep-Signatures/src/utils.py
code served (permissive licence) · get_code("f6b944f50d3f15ae")
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 int, ndim 0, dtype family int, shape () 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)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

ARGS = {
    "model": model
}
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 int, ndim 0, dtype family int, shape () — shape () 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
gen_noise Ran
ran, honoured the contract we drafted
patrick-kidger/Deep-Signature-Transforms/src/generative_model.py
code served (permissive licence) · get_code("934b81ca4ac35740")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned ndarray, ndim 2, dtype family float, shape (2, 50) 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 numpy as np

# Set random seed for reproducibility
np.random.seed(0)

# The function has one parameter: n_points with default value 100
# We can use the default or specify a custom value
# Let's use a small value for testing

ARGS = {
    "n_points": 50
}
Setup block drafted by
a language model, Claude Sonnet 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 ndarray, ndim 2, dtype family float, shape (2, 50) — shape (2, 50) 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
identity Ran
ran, honoured the contract we drafted
patrick-kidger/Deep-Signatures/packages/candle/recurrent.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from dholzmueller/universal_double_descent, is pointer only (licence: NONE) · get_code("9910e2fc297f8665")
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 dholzmueller/universal_double_descent, 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.

loss_fn Ran
ran; our contract draft was wrong, not the code
patrick-kidger/Deep-Signature-Transforms/src/signature_inversion.py
code served (permissive licence) · get_code("56be8c2cedfe1979")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned function. 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

def loss_fn(order):
    normalisation = torch.tensor([np.floor(np.log(i + 1) / np.log(2)) for i in range(1, 2 ** (order + 1) - 1)], dtype=torch.float)

    def loss(output, target):
        output *= normalisation
        target *= normalisation
        return torch.log(((output - target) ** 2).mean())
    return loss

ARGS = {
    'order': 2
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
lyndon_words Ran
ran on a synthesized input
anh-tong/signax/src/signax/utils.py
code served (permissive licence) · get_code("23a544607e61affb")
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. 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 jax
import jax.numpy as jnp

ARGS = {
    "depth": 3,
    "dim": 2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list 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
otimes Ran
ran on a synthesized input
anh-tong/signax/src/signax/tensor_ops.py
code served (permissive licence) · get_code("7a425dc74574f09b")
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 ArrayImpl, ndim 5, dtype family float, shape (2, 2, 2, 2, 2) 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 jax
import jax.numpy as jnp

# Set seed for reproducibility
jax.config.update('jax_default_prng_impl', 'threefry2x32')

# Create small test tensors
# x: 3D tensor of shape (2, 2, 2)
x = jnp.ones((2, 2, 2), dtype=jnp.float32)

# y: 2D tensor of shape (2, 2)
y = jnp.ones((2, 2), dtype=jnp.float32)

ARGS = {
    "x": x,
    "y": y,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ArrayImpl, ndim 5, dtype family float, shape (2, 2, 2, 2, 2) — shape (2, 2, 2, 2, 2) 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
restricted_exp Ran
ran on a synthesized input
anh-tong/signax/src/signax/tensor_ops.py
code served (permissive licence) · get_code("42bf8b2c142070d5")
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. 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 jax
import jax.numpy as jnp

# Create a small input array with shape (n,)
# Using n=4 for a small tensor
input_array = jnp.array([1.0, 2.0, 3.0, 4.0])

# depth parameter - the depth of signature
depth_value = 3

ARGS = {
    "input": input_array,
    "depth": depth_value,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list 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
sig_dim Ran
ran on a synthesized input
patrick-kidger/Deep-Signatures/packages/siglayer/backend.py
code served (permissive licence) · get_code("11d6ef870746b44b")
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 int, ndim 0, dtype family int, shape () 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 numpy as np

# Analyze the function: sig_dim(alphabet_size, depth)
# The function uses:
# - alphabet_size: a numeric value (int or float)
# - depth: a numeric value (int or float)
# Both are used in arithmetic operations

# The function calculates: alphabet_size * (1 - alphabet_size ** depth) / (1 - alphabet_size)
# This is a geometric series formula for signature dimension calculation
# alphabet_size should typically be > 1, and depth should be a positive integer

ARGS = {
    "alphabet_size": 5,
    "depth": 3
}
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 int, ndim 0, dtype family int, shape () — shape () 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
index_select Not yet run
unverified (harvested, no recorded run)
anh-tong/signax/src/signax/utils.py
code served (permissive licence) · get_code("e14367711a5880ab")
How it ran

no invocation recorded.

play Not yet run
unverified (harvested, no recorded run)
patrick-kidger/Deep-Signatures/src/reinforcement_learning.py
code served (permissive licence) · get_code("91d14e4884af0665")
How it ran

no invocation recorded.

train Not yet run
unverified (harvested, no recorded run)
patrick-kidger/Deep-Signatures/src/reinforcement_learning.py
code served (permissive licence) · get_code("987a3ca85150e2e4")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 2 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 2 the same month, 2 later) and 12 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
count_parameters 180 other papers
6 dated earlier, 1 the same month, 173 later · 8 only as an assembly brick
1705.08415 (2017-05)
Supervised Community Detection with Line Graph Neural Networks
1707.01926 (2017-07)
Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting
1808.09367 (2018-08)
Deriving Machine Attention from Human Rationales
and 177 more not named here
identity 69 other papers
5 dated earlier, 1 the same month, 63 later · 5 only as an assembly brick
1706.07296 (2017-06)
A Minimal Developmental Model Can Increase Evolvability in Soft Robots
1802.09419 (2018-02)
Stochastic Hyperparameter Optimization through Hypernetworks
1804.06055 (2018-04)
Co-occurrence Feature Learning from Skeleton Data for Action Recognition and Detection with Hierarchical Aggregation
and 66 more not named here

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

The signature is an infinite graded sequence of statistics known to characterise a stream of data up to a negligible equivalence class. It is a transform which has previously been treated as a fixed feature transformation, on top of which a model may be built. We propose a novel approach which combines the advantages of the signature transform with modern deep learning frameworks. By learning an augmentation of the stream prior to the signature transform, the terms of the signature may be selected in a data-dependent way. More generally, we describe how the signature transform may be used as a layer anywhere within a neural network. In this context it may be interpreted as a pooling operation. We present the results of empirical experiments to back up the theoretical justification. Code available at https://github.com/patrick-kidger/Deep-Signature-Transforms.

For agents

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

get_harvested_code_for_paper("1905.08494")
get_code_for_paper("1905.08494")
have("1905.08494")

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

curl https://syntology.ai/api/ran/1905.08494.json

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

[![Syntology run record](https://syntology.ai/api/ran/1905.08494.svg)](https://syntology.ai/paper/1905.08494)

Connect an agent — have() is free.