SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1803.04831 · 2018

Independently Recurrent Neural Network (IndRNN): Building A Longer and Deeper RNN

arXiv · PDF · Open in the Atlas

Code that ran

5 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 4 have no contract verdict; 1 failure of our instrument; 0 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/1803.04831.json · badge: /api/ran/1803.04831.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
lmnt-com/haste community (classifier label)
classifier basis not recorded on the link
4 ran · 0 no recorded run
StefOe/indrnn-pytorch reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
LayerNormGRUScript Ran
ran on a synthesized input
lmnt-com/haste/frameworks/pytorch/layer_norm_gru.py
code served (permissive licence) · get_code("1f430a28206da9a4")
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 (6, 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.functional as F

torch.manual_seed(0)

# Dimensions: time_steps=5, batch_size=2, hidden_size=4, input_size=3
time_steps = 5
batch_size = 2
hidden_size = 4
input_size = 3

# Build tensors
input_tensor = torch.randn(time_steps, batch_size, input_size)
h0 = torch.randn(batch_size, hidden_size)
kernel = torch.randn(input_size, hidden_size * 3)
recurrent_kernel = torch.randn(hidden_size, hidden_size * 3)
bias = torch.randn(hidden_size * 3)
recurrent_bias = torch.randn(hidden_size * 3)
gamma = torch.randn(2, hidden_size * 3)

# Zoneout mask: shape (time_steps, batch_size, hidden_size)
zoneout_mask = torch.randint(0, 2, (time_steps, batch_size, hidden_size)).float()

ARGS = {
    "training": True,
    "zoneout_prob": 0.5,
    "input": input_tensor,
    "h0": h0,
    "kernel": kernel,
    "recurrent_kernel": recurrent_kernel,
    "bias": bias,
    "recurrent_bias": recurrent_bias,
    "gamma": gamma,
    "zoneout_mask": zoneout_mask,
}
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 3, dtype family float, shape (6, 2, 4) — shape (6, 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
LayerNormIndRNNScript Ran
ran on a synthesized input
lmnt-com/haste/frameworks/pytorch/layer_norm_indrnn.py
code served (permissive licence) · get_code("4c0ac89dfdd49cb8")
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 (4, 2, 5) 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.functional as F

torch.manual_seed(0)

# Determine dimensions
batch_size = 2
time_steps = 3
input_dim = 4
hidden_size = 5

# Build arguments
training = True
zoneout_prob = 0.1

# input: [time_steps, batch_size, input_dim]
input_tensor = torch.randn(time_steps, batch_size, input_dim)

# h0: [batch_size, hidden_size]
h0 = torch.randn(batch_size, hidden_size)

# kernel: [input_dim, hidden_size]
kernel = torch.randn(input_dim, hidden_size)

# recurrent_scale: [hidden_size] or scalar, used in element-wise multiplication
recurrent_scale = torch.randn(hidden_size)

# bias: [hidden_size]
bias = torch.randn(hidden_size)

# gamma: list with at least one element of shape [hidden_size] for layer_norm weight
gamma = [torch.randn(hidden_size)]

# zoneout_mask: [time_steps, batch_size, hidden_size]
zoneout_mask = torch.bernoulli(torch.full((time_steps, batch_size, hidden_size), 1.0 - zoneout_prob))

ARGS = {
    "training": training,
    "zoneout_prob": zoneout_prob,
    "input": input_tensor,
    "h0": h0,
    "kernel": kernel,
    "recurrent_scale": recurrent_scale,
    "bias": bias,
    "gamma": gamma,
    "zoneout_mask": zoneout_mask,
}
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 3, dtype family float, shape (4, 2, 5) — shape (4, 2, 5) 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
LayerNormLSTMScript Ran
ran on a synthesized input
lmnt-com/haste/frameworks/pytorch/layer_norm_lstm.py
code served (permissive licence) · get_code("05bc88c9379472a6")
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 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 torch
import torch.nn.functional as F

torch.manual_seed(0)

# Define dimensions
time_steps = 3
batch_size = 2
input_size = 5
hidden_size = 4

# Construct tensors
input_tensor = torch.randn(time_steps, batch_size, input_size)
h0 = torch.randn(batch_size, hidden_size)
c0 = torch.randn(batch_size, hidden_size)
kernel = torch.randn(input_size, hidden_size * 4)
recurrent_kernel = torch.randn(hidden_size, hidden_size * 4)
bias = torch.randn(hidden_size * 4)
gamma = [torch.ones(hidden_size * 4), torch.ones(hidden_size * 4)]
gamma_h = torch.ones(hidden_size)
beta_h = torch.zeros(hidden_size)
zoneout_mask = torch.bernoulli(torch.full((time_steps, batch_size, 1), 0.5))

ARGS = {
    "training": True,
    "zoneout_prob": 0.1,
    "input": input_tensor,
    "h0": h0,
    "c0": c0,
    "kernel": kernel,
    "recurrent_kernel": recurrent_kernel,
    "bias": bias,
    "gamma": gamma,
    "gamma_h": gamma_h,
    "beta_h": beta_h,
    "zoneout_mask": zoneout_mask,
}
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 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
check_bounds Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
StefOe/indrnn-pytorch/indrnn.py
pointer only (licence: NONE) · get_code("353a35739d84ff7d")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, StefOe/indrnn-pytorch 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.

extract Ran
ran on a synthesized input
lmnt-com/haste/benchmarks/report.py
code served (permissive licence) · get_code("f26710d97fca3491")
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 ndarray, ndim 1, dtype family int, shape (3,) 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

# x should be an array-like that can be filtered
x = np.array([1, 2, 3, 4, 5, 6])

# predicate should be a callable that returns True/False for each element
predicate = lambda elem: elem > 3

ARGS = {
    "x": x,
    "predicate": predicate
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
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 1, dtype family int, shape (3,) — shape (3,) 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

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, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 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.

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

Recurrent neural networks (RNNs) have been widely used for processing sequential data. However, RNNs are commonly difficult to train due to the well-known gradient vanishing and exploding problems and hard to learn long-term patterns. Long short-term memory (LSTM) and gated recurrent unit (GRU) were developed to address these problems, but the use of hyperbolic tangent and the sigmoid action functions results in gradient decay over layers. Consequently, construction of an efficiently trainable deep network is challenging. In addition, all the neurons in an RNN layer are entangled together and their behaviour is hard to interpret. To address these problems, a new type of RNN, referred to as independently recurrent neural network (IndRNN), is proposed in this paper, where neurons in the same layer are independent of each other and they are connected across layers. We have shown that an IndRNN can be easily regulated to prevent the gradient exploding and vanishing problems while allowing the network to learn long-term dependencies. Moreover, an IndRNN can work with non-saturated activation functions such as relu (rectified linear unit) and be still trained robustly. Multiple IndRNNs can be stacked to construct a network that is deeper than the existing RNNs. Experimental results have shown that the proposed IndRNN is able to process very long sequences (over 5000 time steps), can be used to construct very deep networks (21 layers used in the experiment) and still be trained robustly. Better performances have been achieved on various tasks by using IndRNNs compared with the traditional RNN and LSTM. The code is available at https://github.com/Sunnydreamrain/IndRNN_Theano_Lasagne.

For agents

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

get_harvested_code_for_paper("1803.04831")
get_code_for_paper("1803.04831")
have("1803.04831")

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

curl https://syntology.ai/api/ran/1803.04831.json

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

[![Syntology run record](https://syntology.ai/api/ran/1803.04831.svg)](https://syntology.ai/paper/1803.04831)

Connect an agent — have() is free.