SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1607.04606 · Transactions of the Association for Computational Linguistics · 2016

Enriching Word Vectors with Subword Information

arXiv · PDF · Open in the Atlas

Code that ran

25 of 31 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 21 have no contract verdict; 4 failures of our instrument; 6 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/1607.04606.json · badge: /api/ran/1607.04606.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
ulf1/augtxt community (classifier label)
classifier basis not recorded on the link
11 ran · 0 no recorded run
plasticityai/magnitude community (classifier label)
classifier basis not recorded on the link
7 ran · 2 no recorded run
dbaumgarten/FToDTF community (classifier label)
classifier basis not recorded on the link
3 ran · 2 no recorded run
SarangShaikh201/fastText reimplementation (classifier label)
classifier basis: classifier_2025
2 ran · 1 no recorded run
zhang2010hao/cw2vec-pytorch reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
Omerktn/fastText-iterative extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
Kinetikm/fastTextRelearning reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
FunctionStatusWhere it lives
check_glibc_version Ran
ran on a synthesized input
plasticityai/magnitude/glibc.py
code served (permissive licence) · get_code("39f3e4d5f2279ff9")
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 re
import warnings

def check_glibc_version(version_str, required_major, minimum_minor):
    m = re.match('(?P<major>[0-9]+)\\.(?P<minor>[0-9]+)', version_str)
    if not m:
        warnings.warn('Expected glibc version with 2 components major.minor, got: %s' % version_str, RuntimeWarning)
        return False
    return int(m.group('major')) == required_major and int(m.group('minor')) >= minimum_minor

ARGS = {
    'version_str': '2.31',
    'required_major': 2,
    'minimum_minor': 17
}
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 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
create_lookup_tables Ran
ran; our contract draft was wrong, not the code
SarangShaikh201/fastText/utils.py
pointer only (licence: NOASSERTION) · get_code("a16b79c7bc1860e5")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, SarangShaikh201/fastText (NOASSERTION). 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.

draw_index Ran
ran on a synthesized input
ulf1/augtxt/augtxt/typo.py
code served (permissive licence) · get_code("4464b010ea7987c2")
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 numpy as np
import scipy.stats

# Set seed for reproducibility
np.random.seed(42)

# Test case 1: loc as int
ARGS = {
    'n': 10,
    'loc': 5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — 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
drop_word Ran
ran on a synthesized input
ulf1/augtxt/augtxt/order.py
code served (permissive licence) · get_code("839d40e2e8ad110b")
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 str. 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
from typing import List

np.random.seed(0)

ARGS = {
    'original': 'The quick brown fox jumps over the lazy dog',
    'exclude': ['[MASK]', 'the'],
    'punct': '.,;:!?',
    'num_aug': 2
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
entropy Ran
ran on a synthesized input
plasticityai/magnitude/pymagnitude/converter.py
code served (permissive licence) · get_code("6e32ac79f67582c1")
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 float64, ndim 0, 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
from collections import Counter

# Create a counter object with some frequency counts
counter = Counter({
    'bucket_0': 10,
    'bucket_1': 20,
    'bucket_2': 15,
    'bucket_3': 5
})

ARGS = {
    'counter': counter
}
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 float64, ndim 0, dtype family float — no output shape is stated: observe_signature.py recorded float ndim 0, which disagrees with the fixture run
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
fast_md5_file Ran
ran on a synthesized input
plasticityai/magnitude/pymagnitude/converter_shared.py
code served (permissive licence) · get_code("dee6b2dc58e092d3")
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 str. 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 hashlib
import os
import tempfile

# Create a temporary file for testing
with tempfile.NamedTemporaryFile(mode='wb', delete=False, suffix='.bin') as tmp:
    tmp_path = tmp.name
    # Write some test data
    tmp.write(b'test data' * 1000)

try:
    ARGS = {
        'path': tmp_path,
        'block_size': 256 * 128,
        'stream': False,
    }
finally:
    # Note: In actual execution, you'd want to clean up:
    # os.unlink(tmp_path)
    pass
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
find_index Ran
ran on a synthesized input
ulf1/augtxt/augtxt/keyboard_layouts.py
code served (permissive licence) · get_code("8742e7667cc1dafa")
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 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 torch
import numpy as np

# Create a simple keymap dictionary that maps modes to lists of characters
keymap = {
    "us": ["a", "b", "c", "d", "e", "f", "g", "h", "i", "j"],
    "dvorak": ["x", "y", "z", "h", "q", "w", "e", "r", "t", "u"],
    "qwerty": ["q", "w", "e", "r", "t", "y", "u", "i", "o", "p"]
}

# Test character that exists in one of the keymaps
test_char = "h"

ARGS = {
    "c": test_char,
    "keymap": keymap
}
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 1, 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
generate_ngram_per_word Ran
ran on a synthesized input
dbaumgarten/FToDTF/ftodtf/input.py
code served (permissive licence) · get_code("b5b7c0a5146de532")
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.

from nltk import ngrams

ARGS = {
    "word": "hello",
    "ngram_window": 2,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
nltk — 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_darwin_arches Ran
ran on a synthesized input
plasticityai/magnitude/pep425tags.py
code served (permissive licence) · get_code("e3ecbfdb8037b64d")
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.

# Analyze the function to understand what parameters it needs
# get_darwin_arches(major, minor, machine)
# major: int - macOS major version
# minor: int - macOS minor version  
# machine: str - machine architecture string

# The function checks version tuples and architecture strings
# It needs valid integer versions and a valid architecture string

ARGS = {
    "major": 10,
    "minor": 5,
    "machine": "x86_64"
}
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 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_flag Ran
ran on a synthesized input
plasticityai/magnitude/pep425tags.py
code served (permissive licence) · get_code("5066f582560cf77c")
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 sysconfig
import warnings
from unittest.mock import Mock

def mock_fallback():
    """A fallback function that returns a boolean value."""
    return True

ARGS = {
    'var': 'SOABI',
    'fallback': mock_fallback,
    'expected': True,
    'warn': True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
distutils — 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
have_compatible_glibc Ran
ran on a synthesized input
plasticityai/magnitude/glibc.py
code served (permissive licence) · get_code("e0fd811edefc09ba")
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 ctypes
import re
import warnings

# The function have_compatible_glibc takes two parameters:
# - required_major: an integer for the major version number
# - minimum_minor: an integer for the minimum minor version number

# These are simple integer parameters used for version checking.
# We'll provide reasonable test values.

ARGS = {
    "required_major": 2,
    "minimum_minor": 17
}
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 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
load_model Ran
ran; our contract draft was wrong, not the code
zhang2010hao/cw2vec-pytorch/cw2vec.py
pointer only (licence: NONE) · get_code("847e1524ed1bff1a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, zhang2010hao/cw2vec-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.

md5_file Ran
ran on a synthesized input
plasticityai/magnitude/pymagnitude/converter_shared.py
code served (permissive licence) · get_code("a89e48d6397ae9ba")
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 str. 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 tempfile
import os

# Create a temporary file with some content
temp_dir = tempfile.mkdtemp()
temp_file = os.path.join(temp_dir, "test_file.bin")

with open(temp_file, 'wb') as f:
    f.write(b"This is test content for md5 hashing" * 100)

ARGS = {
    "path": temp_file,
    "block_size": 256 * 128
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
merge_words Ran
ran on a synthesized input
ulf1/augtxt/augtxt/punct.py
code served (permissive licence) · get_code("280281e946bcacff")
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 str. 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 copy
import numpy as np

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

# Define the arguments for merge_words function
ARGS = {
    'text_': "Die Bindestrich-Wörter sind da.",
    'sep': [' ', '-', '–'],
    'exclude': ['[MASK]'],
    'num_aug': 1
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
pad_to_length Ran
ran on a synthesized input
dbaumgarten/FToDTF/ftodtf/input.py
code served (permissive licence) · get_code("fa43a85dd995968e")
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 torch

ARGS = {
    'li': ['apple', 'banana', 'cherry'],
    'length': 5,
    'pad': ''
}
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 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
parse_batch_func Ran
ran on a synthesized input
dbaumgarten/FToDTF/ftodtf/model.py
code served (permissive licence) · get_code("00bd40e0432ca4e5")
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 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 tensorflow as tf

ARGS = {
    'batch_size': 2
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
tensorflow — 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
preprocess Ran
ran; our contract draft was wrong, not the code
SarangShaikh201/fastText/utils.py
pointer only (licence: NOASSERTION) · get_code("55d2494d83e5c523")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, SarangShaikh201/fastText (NOASSERTION). 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.

pressed_twice Ran
ran on a synthesized input
ulf1/augtxt/augtxt/typo.py
code served (permissive licence) · get_code("ff4bb2b5311e1044")
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 str. 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 scipy.stats

ARGS = {
    "word": "Hello",
    "loc": "m",
    "keep_case": True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
random_args Ran
ran on a synthesized input
ulf1/augtxt/augtxt/augmenters.py
code served (permissive licence) · get_code("4f5bb4400e32b7e9")
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 torch
import numpy as np
import copy
import scipy.stats

torch.manual_seed(0)

# Analyze the function:
# - random_args takes cfg_ (a dict)
# - It iterates over items in cfg_
# - For each value that is a list or tuple, it randomly picks one element
# - Returns the modified cfg dict

# The function needs a dict with some list/tuple values to demonstrate its functionality

ARGS = {
    "cfg_": {
        "learning_rate": [0.001, 0.01, 0.1],
        "batch_size": [16, 32, 64],
        "model_type": ["lstm", "gru", "transformer"],
        "dropout": [0.2, 0.5],
        "fixed_param": 42,
        "another_fixed": "some_string",
    }
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, torch — 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
read_args Ran
ran; our contract draft was wrong, not the code
Omerktn/fastText-iterative/python/fasttext_module/fasttext/FastText.py
code served (permissive licence) · get_code("f697966cbe929024")
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 tuple, 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.

arg_list = ['value1', 5, True]

arg_dict = {
    'min_count': 10,
    'word_ngrams': 2
}

arg_names = ['input', 'dim', 'verbose', 'minCount', 'wordNgrams', 'lrUpdateRate', 'label', 'pretrainedVectors']

default_values = {
    'dim': 100,
    'verbose': False,
    'minCount': 5,
    'wordNgrams': 1,
    'lrUpdateRate': 100,
    'label': '__label__',
    'pretrainedVectors': ''
}

ARGS = {
    'arg_list': arg_list,
    'arg_dict': arg_dict,
    'arg_names': arg_names,
    'default_values': default_values
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
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 tuple, ndim 1, 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
remove_syntaxinfo Ran
ran on a synthesized input
ulf1/augtxt/augtxt/punct.py
code served (permissive licence) · get_code("37d343258aa5993a")
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 str. 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 re

def remove_syntaxinfo(text: str) -> str:
    """ Remove `.?!;:,` from string (The $. and $, POS tags in STTS)

    Example:
    --------
    import augtxt.punct
    text = ("Die Lehrerin [MASK] einen Roman. "
            "Die Schülerin [MASK] ein Aufsatz, der sehr [MASK] war.")
    augmented = augtxt.punct.remove_punctcomma(text)
    """
    return re.sub('\\s+', ' ', re.sub('[.?!;:,]+', ' ', text)).strip()


ARGS = {
    "text": "Die Lehrerin [MASK] einen Roman. Die Schülerin [MASK] ein Aufsatz, der sehr [MASK] war."
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
swap_consecutive Ran
ran on a synthesized input
ulf1/augtxt/augtxt/order.py
code served (permissive licence) · get_code("2eb54f96aa23c052")
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 str. 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
from typing import List

# Set random seed for reproducibility
np.random.seed(42)

# Define test arguments for swap_consecutive
ARGS = {
    'original': 'The quick brown fox jumps over the lazy dog',
    'exclude': ['[MASK]', 'the'],
    'punct': '.,;:!?',
    'num_aug': 2
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
swap_consecutive Ran
ran on a synthesized input
ulf1/augtxt/augtxt/typo.py
code served (permissive licence) · get_code("48776fc059b48990")
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 str. 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

np.random.seed(seed=42)

ARGS = {
    "word": "Kinder",
    "loc": "middle",
    "keep_case": True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
synonym_replacement Ran
ran on a synthesized input
ulf1/augtxt/augtxt/wordsubs.py
code served (permissive licence) · get_code("e36a7da38a737dd2")
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 scipy.stats
import copy
import itertools

# Define ARGS dictionary with valid arguments for synonym_replacement

ARGS = {
    "original_seqs": [
        ["hello", "world", "test"],
        ["the", "quick", "brown", "fox"]
    ],
    "synonyms": {
        "hello": ["hi", "hey", "greetings"],
        "world": ["earth", "globe"],
        "test": ["exam", "trial"],
        "the": ["a"],
        "quick": ["fast", "swift"],
        "brown": ["tan", "chocolate"],
        "fox": ["canine", "vulpes"]
    },
    "num_augm": 3,
    "min_repl": 1,
    "max_repl": 1.0,
    "keep_case": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
scipy — 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
write_twice Ran
ran on a synthesized input
ulf1/augtxt/augtxt/order.py
code served (permissive licence) · get_code("5f310db16b7971c6")
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 str. 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
from typing import List

# Set random seed for reproducibility
np.random.seed(42)

# Define arguments for write_twice function
ARGS = {
    'original': 'The quick brown fox jumps over the lazy dog',
    'exclude': ['[MASK]', 'the'],
    'punct': '.,;:!?',
    'num_aug': 2
}
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 str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
char_ngrams Not yet run
unverified (harvested, no recorded run)
plasticityai/magnitude/pymagnitude/converter_shared.py
code served (permissive licence) · get_code("478a1a9fc517ce62")
How it ran

no invocation recorded.

compute_similarity Not yet run
unverified (harvested, no recorded run)
Kinetikm/fastTextRelearning/python/doc/examples/train_unsupervised.py
code served (permissive licence) · get_code("411c04ff52be0d9f")
How it ran

no invocation recorded.

compute_word_similarities Not yet run
unverified (harvested, no recorded run)
dbaumgarten/FToDTF/ftodtf/model.py
code served (permissive licence) · get_code("153d1c4c32072113")
How it ran

no invocation recorded.

get_batches Not yet run
unverified (harvested, no recorded run)
SarangShaikh201/fastText/utils.py
pointer only (licence: NOASSERTION) · get_code("cde234836e3f3978")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, SarangShaikh201/fastText (NOASSERTION). 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.

get_config_var Not yet run
unverified (harvested, no recorded run)
plasticityai/magnitude/pep425tags.py
code served (permissive licence) · get_code("f546d4abc08099d7")
How it ran

no invocation recorded.

ngrams_to_vectors Not yet run
unverified (harvested, no recorded run)
dbaumgarten/FToDTF/ftodtf/model.py
code served (permissive licence) · get_code("cf3407d8e4174215")
How it ran

no invocation recorded.

Also present in other papers' code

Of 31 samples lifted from this paper's linked repositories, 31 carry a function identity and 0 do not. Of the 31 distinct functions with an identity, 5 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 2 the same month, 4 later) and 26 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
read_args 3 other papers
1 the same month, 2 later
1607.01759 (2016-07)
Bag of Tricks for Efficient Text Classification
1612.03651 (2016-12)
FastText.zip: Compressing text classification models
2304.04613 (2023-04)
On Evaluation of Bangla Word Analogies
create_lookup_tables 2 other papers
2 later
1612.03651 (2016-12)
FastText.zip: Compressing text classification models
1803.02893 (2018-03)
An efficient framework for learning sentence representations
compute_similarity 1 other paper
1 the same month
1607.01759 (2016-07)
Bag of Tricks for Efficient Text Classification
get_batches 1 other paper
1 later
1612.03651 (2016-12)
FastText.zip: Compressing text classification models
preprocess 1 other paper
1 later
1612.03651 (2016-12)
FastText.zip: Compressing text classification models

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

More are linked than shown here.

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Continuous word representations, trained on large unlabeled corpora are useful for many natural language processing tasks. Popular models that learn such representations ignore the morphology of words, by assigning a distinct vector to each word. This is a limitation, especially for languages with large vocabularies and many rare words. In this paper, we propose a new approach based on the skipgram model, where each word is represented as a bag of character $n$-grams. A vector representation is associated to each character $n$-gram; words being represented as the sum of these representations. Our method is fast, allowing to train models on large corpora quickly and allows us to compute word representations for words that did not appear in the training data. We evaluate our word representations on nine different languages, both on word similarity and analogy tasks. By comparing to recently proposed morphological word representations, we show that our vectors achieve state-of-the-art performance on these tasks.

For agents

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

get_harvested_code_for_paper("1607.04606")
get_code_for_paper("1607.04606")
have("1607.04606")

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

curl https://syntology.ai/api/ran/1607.04606.json

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

[![Syntology run record](https://syntology.ai/api/ran/1607.04606.svg)](https://syntology.ai/paper/1607.04606)

Connect an agent — have() is free.