SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1804.09541 · 2018

QANet: Combining Local Convolution with Global Self-Attention for Reading Comprehension

arXiv · PDF · Open in the Atlas

Code that ran

14 of 19 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; 4 failures of our instrument; 5 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/1804.09541.json · badge: /api/ran/1804.09541.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
BangLiu/QANet-PyTorch community (classifier label)
classifier basis not recorded on the link
5 ran · 0 no recorded run
Tao-Mind/QDD_Net community (classifier label)
classifier basis not recorded on the link
3 ran · 3 no recorded run
andy840314/QANet-pytorch- reimplementation (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
mirbostani/QA-KD-AL extension (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
Francois-Aubet/EQuANt extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
annaorosz/my_qanet_implementation reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
FunctionStatusWhere it lives
OpenAITransformer_loss Ran
ran on a synthesized input
BangLiu/QANet-PyTorch/trainer/loss.py
code served (permissive licence) · get_code("73f7bc81c21d46f4")
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 0, dtype family float, 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)

# Construct loss criteria
lm_criterion = nn.CrossEntropyLoss(reduction='none')
clf_criterion = nn.CrossEntropyLoss(reduction='none')

# Construct tensors
# X shape: (batch_size, num_sequences, seq_length, vocab_dim)
# We need X[:, :, 1:, 0] to work, so last dim must be at least 1
# Let's use: batch=2, sequences=3, seq_length=5, vocab_dim=1
X = torch.randint(0, 10, (2, 3, 5, 1), dtype=torch.long)

# Y shape: (batch_size * num_sequences,) for classification targets
# clf_logits will have shape (batch_size * num_sequences, num_classes)
# So Y should match that
Y = torch.randint(0, 4, (2 * 3,), dtype=torch.long)

# M shape: (batch_size, num_sequences, seq_length)
# Used as: M.view(-1, M.size(2)) and M[:, 1:]
M = torch.ones(2, 3, 5, dtype=torch.float)

# clf_logits shape: (batch_size * num_sequences, num_classes)
# num_classes = 4
clf_logits = torch.randn(2 * 3, 4)

# lm_logits shape must work with:
# lm_criterion(lm_logits, x_shifted) where x_shifted is (batch_size * num_sequences * (seq_length - 1),)
# So lm_logits should be (batch_size * num_sequences * (seq_length - 1), vocab_size)
# vocab_size = 10
lm_logits = torch.randn(2 * 3 * (5 - 1), 10)

# Coefficients
lm_coef = 0.5

ARGS = {
    'lm_criterion': lm_criterion,
    'clf_criterion': clf_criterion,
    'lm_coef': lm_coef,
    'X': X,
    'Y': Y,
    'M': M,
    'clf_logits': clf_logits,
    'lm_logits': lm_logits
}
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 0, dtype family float, 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
PosEncoder Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
andy840314/QANet-pytorch-/models.py
code served (permissive licence) · get_code("5c8c86edd3d3e783")
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 Tensor, ndim 3, dtype family float; no output shape is stated. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is 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 as nn
import math

torch.manual_seed(0)

# Create a small input tensor
# x should be 3D: (batch, channels, length)
# Using distinct dimensions: batch=2, channels=4, length=6
x = torch.randn(2, 4, 6)

# min_timescale and max_timescale are floats with defaults
min_timescale = 1.0
max_timescale = 10000.0

ARGS = {
    'x': x,
    'min_timescale': min_timescale,
    'max_timescale': max_timescale
}
Setup block drafted by
a language model, Claude Sonnet 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 Tensor, ndim 3, dtype family float — no output shape is stated: observe_signature.py could not observe it (call:AssertionError)
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
clones Ran
ran; our contract draft was wrong, not the code
BangLiu/QANet-PyTorch/model/modules/attention.py
pointer only (licence: MIT); the stored copy get_code serves, from smiles724/molformer, is pointer only (licence: NONE) · get_code("cf78c734df121970")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from smiles724/molformer, 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.

convert_idx Ran
ran, honoured the contract we drafted
Francois-Aubet/EQuANt/prepro.py
pointer only (licence: NONE) · get_code("2a6a9ea511d97648")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, Francois-Aubet/EQuANt 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.

convert_idx Ran
ran on a synthesized input
BangLiu/QANet-PyTorch/data_loader/SQuAD.py
code served (permissive licence) · get_code("23587f1e97b0e35d")
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 int; 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

def convert_idx(text, tokens):
    current = 0
    spans = []
    for token in tokens:
        current = text.find(token, current)
        if current < 0:
            print('Token {} cannot be found'.format(token))
            raise Exception()
        spans.append((current, current + len(token)))
        current += len(token)
    return spans

text = "hello world from python"
tokens = ["hello", "world", "from", "python"]

ARGS = {
    "text": text,
    "tokens": tokens
}
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 list, ndim 2, dtype family int — 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
crop_pad Ran
ran on a synthesized input
Tao-Mind/QDD_Net/pre_process.py
code served (permissive licence) · get_code("32abbed1f309ad9f")
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 1, dtype family int; 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)

# max_length: int, the maximum length to crop or pad to
max_length = 10

# word_index: list of integers, representing word indices
word_index = [1, 2, 3, 4, 5]

ARGS = {
    "max_length": max_length,
    "word_index": word_index,
}
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 list, ndim 1, dtype family int — 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
filter_func Ran
ran on a synthesized input
BangLiu/QANet-PyTorch/data_loader/SQuAD.py
code served (permissive licence) · get_code("c4d748327c33e6cc")
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 bool, ndim 0, dtype family float, 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

class Config:
    def __init__(self):
        self.para_limit = 100
        self.ques_limit = 50
        self.ans_limit = 20

config = Config()

example = {
    'context_tokens': ['token'] * 80,
    'ques_tokens': ['token'] * 40,
    'y1s': [5],
    'y2s': [15]
}

ARGS = {
    'config': config,
    'example': example
}
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 bool, ndim 0, dtype family float, 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
get_ids Ran
ran on a synthesized input
Tao-Mind/QDD_Net/pre_process.py
code served (permissive licence) · get_code("fc42ce0856c29480")
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 (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 numpy as np

# The function get_ids takes qids as input
# It iterates through qids, treating each element t_ as a string
# It takes t_[1:] (substring from index 1 onwards) and converts to int
# So qids should be a list/array of strings where each starts with a character followed by digits

qids = ["q123", "q456", "q789", "q999"]

ARGS = {
    "qids": qids
}
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 (4,) — shape (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_texts Ran
ran on a synthesized input
Tao-Mind/QDD_Net/pre_process.py
code served (permissive licence) · get_code("fdbd06a820d2930d")
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 1, 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
import pandas as pd
import tempfile
import os

# Create temporary CSV files for testing
temp_dir = tempfile.mkdtemp()

# Create question_path CSV with words column
# q0, q1, q2, q3, etc. as indices
questions_data = {
    'words': [
        'hello world',
        'what is',
        'how are',
        'python code',
        'data science'
    ]
}
questions_df = pd.DataFrame(questions_data)
question_path = os.path.join(temp_dir, 'questions.csv')
questions_df.to_csv(question_path, index=False)

# Create file_path CSV with q1 and q2 columns containing question IDs
file_data = {
    'q1': ['q0', 'q1', 'q2'],
    'q2': ['q1', 'q2', 'q3']
}
file_df = pd.DataFrame(file_data)
file_path = os.path.join(temp_dir, 'pairs.csv')
file_df.to_csv(file_path, index=False)

ARGS = {
    'file_path': file_path,
    'question_path': question_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, 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
get_timing_signal Ran
ran, honoured the contract we drafted
mirbostani/QA-KD-AL/qa/model/qanet.py
code served (permissive licence) · get_code("f82361ad7330d575")
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 Tensor, ndim 3, dtype family float, shape (1, 8, 6) 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 math

torch.manual_seed(0)

ARGS = {
    'length': 8,
    'channels': 6,
    'min_timescale': 1.0,
    'max_timescale': 10000.0
}
Setup block drafted by
a language model, Claude Sonnet 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 Tensor, ndim 3, dtype family float, shape (1, 8, 6) — shape (1, 8, 6) 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_timing_signal Ran
ran, honoured the contract we drafted
andy840314/QANet-pytorch-/models.py
code served (permissive licence) · get_code("d2e8e18023f286ab")
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 Tensor, ndim 3, dtype family float, shape (1, 8, 6) 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
import math

torch.manual_seed(0)

# The function takes length (int), channels (int), and two optional float parameters
# length: sequence length
# channels: number of channels (embedding dimension)
# min_timescale: minimum timescale (default 1.0)
# max_timescale: maximum timescale (default 10000.0)

ARGS = {
    'length': 8,
    'channels': 6,
    'min_timescale': 1.0,
    'max_timescale': 10000.0
}
Setup block drafted by
a language model, Claude Sonnet 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 Tensor, ndim 3, dtype family float, shape (1, 8, 6) — shape (1, 8, 6) 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
mask_logits Ran
ran; our contract draft was wrong, not the code
mirbostani/QA-KD-AL/qa/model/qanet.py
code served (permissive licence) · get_code("9dc49f46edd7a69d")
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 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)

# Create target tensor (logits) - shape (2, 4)
target = torch.randn(2, 4)

# Create mask tensor - shape (2, 4) with boolean or integer values
# Using integers 0 and 1 to represent mask
mask = torch.tensor([[1, 1, 0, 1],
                     [0, 1, 1, 0]])

ARGS = {
    "target": target,
    "mask": mask
}
Setup block drafted by
a language model, Claude Sonnet 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 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
mask_logits Ran
ran; our contract draft was wrong, not the code
andy840314/QANet-pytorch-/models.py
code served (permissive licence) · get_code("dea46bb8718e5def")
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 Tensor, ndim 2, dtype family float, shape (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

torch.manual_seed(0)

# Create input tensors with distinct dimensions
# inputs: a 2D tensor representing logits (batch_size=2, seq_len=5)
inputs = torch.randn(2, 5)

# mask: a boolean or integer tensor with same shape as inputs
# where 1 means keep and 0 means mask out
mask = torch.tensor([[1, 1, 0, 1, 0],
                     [1, 0, 0, 1, 1]], dtype=torch.int32)

ARGS = {
    'inputs': inputs,
    'mask': mask
}
Setup block drafted by
a language model, Claude Sonnet 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 Tensor, ndim 2, dtype family float, shape (2, 5) — shape (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
my_loss Ran
ran on a synthesized input
BangLiu/QANet-PyTorch/trainer/loss.py
code served (permissive licence) · get_code("cb4ce684e3929e65")
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 0, dtype family float, 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.functional as f

torch.manual_seed(0)

# y_input should be log probabilities (output of log_softmax)
# Shape: (batch_size, num_classes)
y_input = f.log_softmax(torch.randn(2, 4), dim=1)

# y_target should be class indices
# Shape: (batch_size,)
y_target = torch.tensor([0, 3])

ARGS = {
    "y_input": y_input,
    "y_target": y_target,
}
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 0, dtype family float, 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
PosEncoder Not yet run
unverified (harvested, no recorded run)
mirbostani/QA-KD-AL/qa/model/qanet.py
code served (permissive licence) · get_code("ee4ade0d8bcf356a")
How it ran

no invocation recorded.

add_dense_layer Not yet run
unverified (harvested, no recorded run)
Tao-Mind/QDD_Net/models/layers.py
code served (permissive licence) · get_code("edf10b1abda15159")
How it ran

no invocation recorded.

biGRUs Not yet run
unverified (harvested, no recorded run)
Tao-Mind/QDD_Net/models/layers.py
code served (permissive licence) · get_code("d6c2afca4b825f12")
How it ran

no invocation recorded.

dot_attention Not yet run
unverified (harvested, no recorded run)
Tao-Mind/QDD_Net/models/layers.py
code served (permissive licence) · get_code("96be8bbb7b022500")
How it ran

no invocation recorded.

highway Not yet run
unverified (harvested, no recorded run)
annaorosz/my_qanet_implementation/QANet_keras.py
pointer only (licence: NONE) · get_code("9549014140b95768")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, annaorosz/my_qanet_implementation 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. No invocation recorded.

Also present in other papers' code

Of 19 samples lifted from this paper's linked repositories, 19 carry a function identity and 0 do not. Of the 18 distinct functions with an identity, 5 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 4 later) and 13 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
clones 79 other papers
4 dated earlier, 75 later · 11 only as an assembly brick
1411.4555 (2014-11)
Show and Tell: A Neural Image Caption Generator
1612.00563 (2016-12)
Self-critical Sequence Training for Image Captioning
1706.03762 (2017-06)
Attention Is All You Need
and 76 more not named here
PosEncoder 1 other paper
1 later
2109.12662 (2021-09)
Improving Question Answering Performance Using Knowledge Distillation and Active Learning
convert_idx 1 other paper
1 dated earlier
1508.06615 (2015-08)
Character-Aware Neural Language Models
get_timing_signal 1 other paper
1 later
2109.12662 (2021-09)
Improving Question Answering Performance Using Knowledge Distillation and Active Learning
mask_logits 1 other paper
1 later
2109.12662 (2021-09)
Improving Question Answering Performance Using Knowledge Distillation and Active Learning

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

Current end-to-end machine reading and question answering (Q\&amp;A) models are primarily based on recurrent neural networks (RNNs) with attention. Despite their success, these models are often slow for both training and inference due to the sequential nature of RNNs. We propose a new Q\&amp;A architecture called QANet, which does not require recurrent networks: Its encoder consists exclusively of convolution and self-attention, where convolution models local interactions and self-attention models global interactions. On the SQuAD dataset, our model is 3x to 13x faster in training and 4x to 9x faster in inference, while achieving equivalent accuracy to recurrent models. The speed-up gain allows us to train the model with much more data. We hence combine our model with data generated by backtranslation from a neural machine translation model. On the SQuAD dataset, our single model, trained with augmented data, achieves 84.6 F1 score on the test set, which is significantly better than the best published F1 score of 81.8.

For agents

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

get_harvested_code_for_paper("1804.09541")
get_code_for_paper("1804.09541")
have("1804.09541")

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

curl https://syntology.ai/api/ran/1804.09541.json

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

[![Syntology run record](https://syntology.ai/api/ran/1804.09541.svg)](https://syntology.ai/paper/1804.09541)

Connect an agent — have() is free.