SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2003.04297 · 2020

Improved Baselines with Momentum Contrastive Learning

Xinlei Haoqi, Fan Ross, Girshick Kaiming

arXiv · PDF · Open in the Atlas

Code that ran

28 of 43 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, 1 violated it and 19 have no contract verdict; 8 failures of our instrument; 15 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/2003.04297.json · badge: /api/ran/2003.04297.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
derrickwang005/mosrep community (classifier label)
classifier basis not recorded on the link
10 ran · 1 no recorded run
lavoiems/SADt community (classifier label)
classifier basis not recorded on the link
8 ran · 8 no recorded run
PaperCodeReview/MoCo-TF community (classifier label)
classifier basis not recorded on the link
5 ran · 0 no recorded run
copy not recorded community
not in the paper's declared repositories
3 ran · 3 no recorded run
CoinCheung/denseCL reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
vturrisi/solo-learn extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
1170500804/MoCo reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 3 no recorded run
FunctionStatusWhere it lives
accuracy Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
this paper's copy was not recorded; identical code first harvested from joshr17/IFM
pointer only · get_code("983576fa67f4f01e")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

accuracy Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
this paper's copy was not recorded; identical code first harvested from songweige/contrastive-learning-with-non-semantic-negatives
pointer only · get_code("131a82fd65128218")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

accuracy Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
CoinCheung/denseCL/main_densecl.py
pointer only (licence: NOASSERTION) · get_code("06532ecb66c0bcf9")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, CoinCheung/denseCL (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.

accuracy Ran
ran on a synthesized input
derrickwang005/mosrep/utils/misc.py
code served (permissive licence) · get_code("c877664f0b6c29ba")
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

torch.manual_seed(0)

# Create small tensors for the accuracy function
batch_size = 2
num_classes = 4

# output: tensor of logits/scores, shape (batch_size, num_classes)
output = torch.randn(batch_size, num_classes)

# target: tensor of class indices, shape (batch_size,)
target = torch.tensor([0, 2])

# topk: tuple of k values to compute accuracy for
topk = (1, 2)

ARGS = {
    "output": output,
    "target": target,
    "topk": topk,
}
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
check_arguments Ran
ran on a synthesized input
PaperCodeReview/MoCo-TF/common.py
code served (permissive licence) · get_code("129adedbc3f353c4")
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 Namespace. 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
from argparse import Namespace

# Create a simple namespace object with the required attributes
args = Namespace(
    src_path='/tmp/src',
    data_path='/tmp/data',
    result_path='/tmp/result'
)

ARGS = {'args': args}
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 Namespace — a Namespace output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Namespace 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
conv1x1 Ran
ran; our contract draft was wrong, not the code
derrickwang005/mosrep/moco/resnet.py
pointer only (licence: MIT); the stored copy get_code serves, from medaitech/scrd4ad, is pointer only (licence: NONE) · get_code("2a80220dabcb742a")
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 medaitech/scrd4ad, 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.

conv3x3 Ran
ran; our contract draft was wrong, not the code
derrickwang005/mosrep/moco/resnet.py
pointer only (licence: MIT); the stored copy get_code serves, from medaitech/scrd4ad, is pointer only (licence: NONE) · get_code("600ff2c45e0de056")
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 medaitech/scrd4ad, 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.

get_args Ran
ran on a synthesized input
lavoiems/SADt/src/common/util.py
code served (permissive licence) · get_code("b615a4c82ed53c29")
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 os
import json
import tempfile

# Create a temporary directory and args.json file
temp_dir = tempfile.mkdtemp()
args_file_path = os.path.join(temp_dir, 'args.json')

# Create a sample args.json file
sample_args = {
    "learning_rate": 0.001,
    "batch_size": 32,
    "epochs": 10,
    "model_name": "test_model"
}

with open(args_file_path, 'w') as f:
    json.dump(sample_args, f)

ARGS = {
    "save_path": temp_dir
}
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 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
get_args_dict Ran
ran on a synthesized input
lavoiems/SADt/src/common/util.py
code served (permissive licence) · get_code("3a56153874d5df28")
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 torch.nn as nn

class MockArgs:
    def __init__(self):
        self.learning_rate = 0.001
        self.batch_size = 32
        self.epochs = 10
        self.model_name = "test_model"
        self.use_cuda = False
        self.dropout = 0.5
        self.hidden_dim = 128
        self.vocab_size = 1000
        self.max_length = 512
        self.seed = 42
        self.num_layers = 2
        self.custom_dict = {"key": "value"}
        self.custom_list = [1, 2, 3]
        self.custom_tuple = (1, 2, 3)
        self.threshold = 0.7
        self.is_debug = True
        self.count = 5
        # Non-builtin types that should be filtered out
        self.tensor = torch.randn(2, 4)
        self.module = nn.Linear(4, 3)

args = MockArgs()

ARGS = {"args": args}
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 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
get_caller_name Ran
ran on a synthesized input
derrickwang005/mosrep/utils/misc.py
code served (permissive licence) · get_code("ada42cd4264e001d")
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 inspect

def get_caller_name(depth=0):
    """
    Args:
        depth (int): Depth of caller conext, use 0 for caller depth.
        Default value: 0.

    Returns:
        str: module name of the caller
    """
    frame = inspect.currentframe().f_back
    for _ in range(depth):
        frame = frame.f_back
    return frame.f_globals['__name__']


ARGS = {
    'depth': 0
}
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
get_linear_transform Ran
ran on a synthesized input
derrickwang005/mosrep/utils/dataset.py
code served (permissive licence) · get_code("5636af1506db32d1")
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 Compose. 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 torchvision.transforms as transforms

torch.manual_seed(0)

ARGS = {
    "args": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Compose — a Compose output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Compose 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
get_local_transform Ran
ran on a synthesized input
derrickwang005/mosrep/utils/dataset.py
code served (permissive licence) · get_code("f8f9437c5a9c0588")
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 Compose. 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 random
import torch
import torchvision.transforms as transforms
from PIL import Image, ImageFilter, ImageOps

torch.manual_seed(0)
random.seed(0)

class ArgsConfig:
    def __init__(self):
        self.local_size = 224
        self.local_scale = (0.08, 1.0)

ARGS = {
    "args": ArgsConfig()
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Compose — a Compose output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Compose 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
get_logger Ran
ran on a synthesized input
PaperCodeReview/MoCo-TF/common.py
code served (permissive licence) · get_code("e9cdcc13fe49820c")
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 Logger. 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 sys
import logging

# The function get_logger takes a single parameter: name (a string)
# It uses this name to create a logger via logging.getLogger(name)
# No complex objects needed - just a simple string

ARGS = {
    "name": "test_logger"
}
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 Logger — a Logger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Logger 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
get_standard_transform Ran
ran on a synthesized input
derrickwang005/mosrep/utils/dataset.py
code served (permissive licence) · get_code("045e7fb3560388d4")
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 Compose. 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 random
import torch
import torchvision.transforms as transforms
from PIL import Image

torch.manual_seed(0)
random.seed(0)

class ArgsConfig:
    def __init__(self):
        self.global_size = 224
        self.global_scale = (0.08, 1.0)

ARGS = {
    "args": ArgsConfig()
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Compose — a Compose output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Compose 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
has_models Ran
ran on a synthesized input
lavoiems/SADt/src/common/initialize.py
code served (permissive licence) · get_code("2da6bd8937cdb982")
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 os
import tempfile
import shutil

# Create a temporary directory with some files in it
temp_dir = tempfile.mkdtemp()
# Add a dummy file to make the directory non-empty
with open(os.path.join(temp_dir, "model.txt"), "w") as f:
    f.write("dummy")

ARGS = {
    "path": temp_dir
}
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
infer_iteration Ran
ran on a synthesized input
lavoiems/SADt/src/common/initialize.py
code served (permissive licence) · get_code("60ae2cf6b5924c2d")
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 os
import tempfile
import shutil

# Create temporary directories for testing
temp_dir = tempfile.mkdtemp()
model_path = os.path.join(temp_dir, 'model')
save_path = temp_dir

os.makedirs(model_path, exist_ok=True)

# Create some dummy model files to make has_models return True
# Files named like "model_name:0.pt", "model_name:1.pt", etc.
with open(os.path.join(model_path, 'test_model:0.pt'), 'w') as f:
    f.write('dummy')
with open(os.path.join(model_path, 'test_model:1.pt'), 'w') as f:
    f.write('dummy')
with open(os.path.join(model_path, 'test_model:5.pt'), 'w') as f:
    f.write('dummy')

ARGS = {
    'name': 'test_model',
    'reload': True,
    'model_path': model_path,
    'save_path': save_path
}
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 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
initialize Ran
ran on a synthesized input
lavoiems/SADt/src/common/initialize.py
code served (permissive licence) · get_code("5f912135342a1eac")
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 os
import torch
import torch.nn as nn
import tempfile
import shutil

torch.manual_seed(0)

# Create a minimal nn.Module for testing
class MinimalModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

# Create temporary directory structure
temp_dir = tempfile.mkdtemp()
model_dir = os.path.join(temp_dir, 'model')
os.makedirs(model_dir, exist_ok=True)

# Create dummy model files with proper naming format (name:id.pt)
model1 = MinimalModel()
model2 = MinimalModel()
model3 = MinimalModel()

torch.save(model1.state_dict(), os.path.join(model_dir, 'model_a:1.pt'))
torch.save(model2.state_dict(), os.path.join(model_dir, 'model_a:2.pt'))
torch.save(model3.state_dict(), os.path.join(model_dir, 'model_b:1.pt'))

# Build models dict with fresh instances to load into
models = {
    'model_a': MinimalModel(),
    'model_b': MinimalModel(),
}

ARGS = {
    'models': models,
    'reload': True,
    'dir': temp_dir,
    'model_path': model_dir,
}
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 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
is_global_master Ran
ran, violated the contract we drafted
derrickwang005/mosrep/utils/misc.py
pointer only (licence: MIT); the stored copy get_code serves, from tileb1/simzss, is pointer only (licence: NONE) · get_code("7c29b374198bddcd")
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 tileb1/simzss, 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.

mocov2plus_loss_func Ran
ran; our contract draft was wrong, not the code
vturrisi/solo-learn/solo/losses/mocov2plus.py
code served (permissive licence) · get_code("4dd7c3cff0150cb9")
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 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)

# Create query: NxD tensor (batch_size=2, embedding_dim=4)
query = torch.randn(2, 4)
query = F.normalize(query, dim=1)  # Normalize as typical in contrastive learning

# Create key: NxD tensor (batch_size=2, embedding_dim=4)
key = torch.randn(2, 4)
key = F.normalize(key, dim=1)

# Create queue: DxK tensor (embedding_dim=4, queue_size=6)
# Note: queue is transposed (D x K) for efficient einsum computation
queue = torch.randn(4, 6)
queue = F.normalize(queue, dim=0)  # Normalize along embedding dimension

# Temperature parameter
temperature = 0.1

ARGS = {
    'query': query,
    'key': key,
    'queue': queue,
    'temperature': temperature
}
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 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
normalize Ran
ran on a synthesized input
lavoiems/SADt/src/common/util.py
code served (permissive licence) · get_code("c3f9a96837f42be0")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function normalizes a tensor by:
# 1. Clamping values to [-1, 1]
# 2. Adding 1 (shifting to [0, 2])
# 3. Multiplying by 0.5 (scaling to [0, 1])
# It modifies the tensor in-place and returns it

x = torch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
one_hot_embedding Ran
ran on a synthesized input
lavoiems/SADt/src/models/EGSC/model.py
code served (permissive licence) · get_code("1941df4f816178f1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (4, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Construct arguments for one_hot_embedding
labels = torch.tensor([0, 1, 0, 2], dtype=torch.long)
num_classes = 4

ARGS = {
    "labels": labels,
    "num_classes": num_classes,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (4, 4) — shape (4, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
readfile Ran
ran on a synthesized input
derrickwang005/mosrep/imagenet2wds.py
code served (permissive licence) · get_code("efba4631d8880f62")
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 bytes. 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 binary 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"test binary content")

ARGS = {
    "fname": temp_file
}
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 bytes — a bytes output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a bytes 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
resnet18 Ran
ran; our contract draft was wrong, not the code
derrickwang005/mosrep/moco/resnet.py
code served (permissive licence) · get_code("580cd13eb9675631")
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 ResNet. 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 torch.nn as nn
from typing import List, Optional, Callable

torch.manual_seed(0)

def resnet18(pretrained: bool = False, progress: bool = True, **kwargs) -> nn.Module:
    """Constructs a ResNet-18 model."""
    return nn.Module()

ARGS = {
    'pretrained': False,
    'progress': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torchvision — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ResNet — a ResNet output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ResNet 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
search_same Ran
ran on a synthesized input
PaperCodeReview/MoCo-TF/common.py
code served (permissive licence) · get_code("5c153d8093022479")
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 os
import yaml
import numpy as np
import pandas as pd
import tempfile
import shutil

# Create a temporary directory structure for testing
temp_dir = tempfile.mkdtemp()
result_path = temp_dir
task = "test_task"

# Create necessary directory structure
task_dir = os.path.join(result_path, task)
os.makedirs(task_dir, exist_ok=True)

# Create a stamp directory with model_desc.yml and checkpoint/history
stamp = "stamp_001"
stamp_dir = os.path.join(task_dir, stamp)
os.makedirs(stamp_dir, exist_ok=True)

checkpoint_dir = os.path.join(stamp_dir, "checkpoint")
os.makedirs(checkpoint_dir, exist_ok=True)

history_dir = os.path.join(stamp_dir, "history")
os.makedirs(history_dir, exist_ok=True)

# Create model_desc.yml
model_desc = {
    'task': task,
    'epochs': 10,
    'learning_rate': 0.001,
    'batch_size': 32,
    'model': 'test_model',
}
with open(os.path.join(stamp_dir, "model_desc.yml"), 'w') as f:
    yaml.dump(model_desc, f)

# Create history/epoch.csv with some training data
df = pd.DataFrame({
    'epoch': [0, 1, 2],
    'loss': [1.5, 1.2, 0.9],
    'accuracy': [0.6, 0.7, 0.8]
})
df.to_csv(os.path.join(history_dir, "epoch.csv"), index=False)

# Create a checkpoint file
checkpoint_path = os.path.join(checkpoint_dir, "003_checkpoint.index")
with open(checkpoint_path, 'w') as f:
    f.write("dummy checkpoint")

# Create mock args object
class Args:
    def __init__(self):
        self.result_path = result_path
        self.task = task
        self.epochs = 10
        self.learning_rate = 0.001
        self.batch_size = 32
        self.model = 'test_model'
        self.checkpoint = None
        self.history = None
        self.tensorboard = None
        self.tb_interval = 100
        self.snapshot = None
        self.summary = None
        self.src_path = None
        self.data_path = None
        self.resume = False
        self.stamp = None
        self.gpus = "0"
        self.ignore_search = ""

args = Args()

ARGS = {
    'args': args
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, pandas, yaml — 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
set_dataset Ran
ran on a synthesized input
PaperCodeReview/MoCo-TF/dataloader.py
code served (permissive licence) · get_code("1d3bcae9b7aa7920")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 3, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
import numpy as np
import pandas as pd

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

# Create imagenet_trainset.csv with sample data
trainset_data = {
    'path': ['image1.jpg', 'image2.jpg'],
    'label': [0, 1]
}
trainset_df = pd.DataFrame(trainset_data)
trainset_df.to_csv(os.path.join(temp_dir, 'imagenet_trainset.csv'), index=False)

# Create imagenet_valset.csv with sample data
valset_data = {
    'path': ['val_image1.jpg', 'val_image2.jpg'],
    'label': [0, 1]
}
valset_df = pd.DataFrame(valset_data)
valset_df.to_csv(os.path.join(temp_dir, 'imagenet_valset.csv'), index=False)

# Define ARGS for the function
ARGS = {
    'task': 'lincls',
    'data_path': temp_dir
}
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 tuple, ndim 3, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
set_lincls Ran
ran on a synthesized input
PaperCodeReview/MoCo-TF/model.py
code served (permissive licence) · get_code("d444d3b77bdf398b")
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 Functional. 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
from tensorflow.keras.layers import Dense, Input, GlobalAveragePooling2D
from tensorflow.keras import Model
from tensorflow.keras.regularizers import l2


class Args:
    def __init__(self):
        self.use_bias = True
        self.weight_decay = 0.0001
        self.freeze = False
        self.classes = 10


def create_mock_backbone():
    inputs = Input(shape=(8, 8, 3))
    x = GlobalAveragePooling2D(name='avg_pool')(inputs)
    backbone = Model(inputs=inputs, outputs=x, name='backbone')
    return backbone


args = Args()
backbone = create_mock_backbone()

ARGS = {
    'args': args,
    'backbone': backbone
}
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 Functional — a Functional output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Functional 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
triple_channel Ran
ran on a synthesized input
lavoiems/SADt/src/common/loaders/images.py
code served (permissive licence) · get_code("8685ce39337a0ffc")
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 (3, 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)

# The function takes a tensor x and either returns it as-is if it has 3 channels,
# or concatenates it 3 times along dimension 0 to create 3 channels.

# Test case 1: tensor with 3 channels already (should return as-is)
x_3channel = torch.randn(3, 2, 4)

# Test case 2: tensor with 1 channel (should be tripled to 3)
x_1channel = torch.randn(1, 2, 4)

# Use the 1-channel case as it exercises the main logic
ARGS = {
    "x": x_1channel
}
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 (3, 2, 4) — shape (3, 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
validate Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
this paper's copy was not recorded; identical code first harvested from songweige/contrastive-learning-with-non-semantic-negatives
pointer only · get_code("63ce1d5edf8fb0a0")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

accuracy Not yet run
unverified (harvested, no recorded run)
this paper's copy was not recorded; identical code first harvested from zbh2047/clipping-algorithms
pointer only; the stored copy get_code serves, from zbh2047/clipping-algorithms, is pointer only (licence: NONE) · get_code("9b8289076669fe4f")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

build_model Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/models/EGSC/model.py
code served (permissive licence) · get_code("3b69089e7dfcef76")
How it ran

no invocation recorded.

build_model Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/models/sg/model.py
code served (permissive licence) · get_code("82c869e73df33e11")
How it ran

no invocation recorded.

calculate_activation_statistics Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/common/evaluation/fid.py
code served (permissive licence) · get_code("e31ff5999b5f7a9f")
How it ran

no invocation recorded.

calculate_frechet_distance Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/common/evaluation/fid.py
code served (permissive licence) · get_code("0def50a351111624")
How it ran

no invocation recorded. this sample carries JSON test cases instead of a setup block; get_reference_implementation serves them with the level they back.

concat_all_gather Not yet run
unverified (harvested, no recorded run)
this paper's copy was not recorded; identical code first harvested from sail-sg/iFormer
pointer only · get_code("73cecca9f3575f09")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

concat_all_gather Not yet run
unverified (harvested, no recorded run)
derrickwang005/mosrep/moco/builder.py
pointer only (licence: MIT); the stored copy get_code serves, from tntek/source-free-domain-adaptation, is pointer only (licence: NONE) · get_code("a2ae53d34bc13317")
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 tntek/source-free-domain-adaptation, 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. No invocation recorded.

construct_confusion_matrix_image Not yet run
unverified (harvested, no recorded run)
1170500804/MoCo/main_lincls.py
pointer only (licence: NOASSERTION) · get_code("c6c33546958d1759")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, 1170500804/MoCo (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_activations Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/common/evaluation/fid.py
code served (permissive licence) · get_code("96aec64c84c32d81")
How it ran

no invocation recorded.

parse_args Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/evaluate.py
pointer only (licence: MIT); the stored copy get_code serves, from lavoiems/Cats-UDT, is pointer only (licence: NONE) · get_code("c26e2d748dacb750")
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 lavoiems/Cats-UDT, 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. No invocation recorded.

svhn Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/common/loaders/images.py
code served (permissive licence) · get_code("35ee0a5ef1232571")
How it ran

no invocation recorded.

svhn_extra Not yet run
unverified (harvested, no recorded run)
lavoiems/SADt/src/common/loaders/images.py
code served (permissive licence) · get_code("07f548c71e7969f6")
How it ran

no invocation recorded.

train Not yet run
unverified (harvested, no recorded run)
1170500804/MoCo/main_lincls.py
pointer only (licence: NOASSERTION) · get_code("efad379ea20fec59")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, 1170500804/MoCo (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.

validate Not yet run
unverified (harvested, no recorded run)
this paper's copy was not recorded; identical code first harvested from ssnl/moco_align_uniform
pointer only; the stored copy get_code serves, from ssnl/moco_align_uniform, is pointer only (licence: NOASSERTION) · get_code("0358792e34e9cbb2")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. 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.

validate Not yet run
unverified (harvested, no recorded run)
1170500804/MoCo/main_lincls.py
pointer only (licence: NOASSERTION) · get_code("3d39d5577a13c30e")
How it ran

The body of this sample is withheld: GitHub could not classify the licence of this paper's repository, 1170500804/MoCo (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.

Also present in other papers' code

Of 43 samples lifted from this paper's linked repositories, 43 carry a function identity and 0 do not. Of the 42 distinct functions with an identity, 19 are also present in other papers whose code we hold (14 with at least one paper dated earlier, 4 the same month, 15 later) and 23 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
concat_all_gather 138 other papers
5 dated earlier, 133 later · 33 only as an assembly brick
1710.09412 (2017-10)
mixup: BEYOND EMPIRICAL RISK MINIMIZATION
1905.04899 (2019-05)
CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features
1905.12794 (2019-05)
Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback
and 135 more not named here
accuracy 116 other papers
6 dated earlier, 1 the same month, 109 later · 8 only as an assembly brick
1502.03044 (2015-02)
Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
1610.05820 (2016-10)
Membership Inference Attacks Against Machine Learning Models
1905.03696 (2019-05)
HAWQ: Hessian AWare Quantization of Neural Networks with Mixed-Precision
and 113 more not named here
conv1x1 112 other papers
5 dated earlier, 107 later · 6 only as an assembly brick
1512.03385 (2015-12)
Deep Residual Learning for Image Recognition
1806.01531 (2018-06)
Deep Mixture of Experts via Shallow Embedding
1905.12794 (2019-05)
Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback
and 109 more not named here
conv3x3 111 other papers
4 dated earlier, 107 later · 6 only as an assembly brick
1512.03385 (2015-12)
Deep Residual Learning for Image Recognition
1806.01531 (2018-06)
Deep Mixture of Experts via Shallow Embedding
1905.12794 (2019-05)
Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback
and 108 more not named here
accuracy 80 other papers
30 dated earlier, 50 later
1307.0414 (2013-07)
Challenges in Representation Learning: A report on three machine learning contests
1412.6553 (2014-12)
Speeding-up Convolutional Neural Networks Using Fine-tuned CP-Decomposition
1503.02531 (2015-03)
Distilling the Knowledge in a Neural Network
and 77 more not named here
calculate_frechet_distance 74 other papers
10 dated earlier, 2 the same month, 62 later
1610.09585 (2016-10)
Conditional Image Synthesis With Auxiliary Classifier GANs
1706.08500 (2017-06)
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
1807.11458 (2018-07)
To learn image super-resolution, use a GAN to learn how to do image degradation first
and 71 more not named here
resnet18 36 other papers
2 dated earlier, 34 later
1905.12794 (2019-05)
Fashion IQ: A New Dataset Towards Retrieving Images by Natural Language Feedback
2002.09564 (2020-02)
Towards Robust and Reproducible Active Learning Using Neural Networks
2103.13859 (2021-03)
Group-CAM: Group Score-Weighted Visual Explanations for Deep Convolutional Networks
and 33 more not named here
accuracy 17 other papers
2 dated earlier, 15 later
1805.03438 (2018-05)
Robust Classification with Convolutional Prototype Learning
1905.11926 (2019-05)
Network Deconvolution
2104.02057 (2021-04)
An Empirical Study of Training Self-Supervised Vision Transformers
and 14 more not named here
validate 14 other papers
2 dated earlier, 12 later
1706.02677 (2017-06)
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
1911.12207 (2019-11)
Orthogonal Convolutional Neural Networks
2005.09681 (2020-05)
Weakly Supervised Representation Learning with Coarse Labels
and 11 more not named here
is_global_master 10 other papers
10 later
2302.14042 (2023-02)
Knowledge-enhanced Visual-Language Pre-training on Chest Radiology Images
2308.05734 (2023-08)
AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining
2311.17048 (2023-11)
Zero-shot Referring Expression Comprehension via Structural Similarity Between Images and Captions
and 7 more not named here
concat_all_gather 5 other papers
1 dated earlier, 4 later · 2 only as an assembly brick
1904.06487 (2019-04)
Semi-supervised Domain Adaptation via Minimax Entropy
2204.10377 (2022-04)
Contrastive Test-Time Adaptation
2206.15369 (2022-06)
No Reason for No Supervision: Improved Generalization in Supervised Models
and 2 more not named here
mocov2plus_loss_func 5 other papers
5 later
2007.06346 (2020-07)
Whitening for Self-Supervised Representation Learning
2403.12448 (2024-03)
Do Generated Data Always Help Contrastive Learning?
2403.12459 (2024-03)
Non-negative Contrastive Learning
and 2 more not named here
accuracy 2 other papers
1 dated earlier, 1 later
1706.02677 (2017-06)
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
2011.09157 (2020-11)
Dense Contrastive Learning for Self-Supervised Visual Pre-Training
get_logger 2 other papers
1 the same month, 1 later
2003.13678 (2020-03)
Designing Network Design Spaces
2004.11362 (2020-04)
Supervised Contrastive Learning
check_arguments 1 other paper
1 the same month
2003.13678 (2020-03)
Designing Network Design Spaces
construct_confusion_matrix_image 1 other paper
1 dated earlier
1706.02677 (2017-06)
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
parse_args 1 other paper
1 later
2010.01262 (2020-10)
Integrating Categorical Semantics into Unsupervised Domain Translation
train 1 other paper
1 dated earlier
1706.02677 (2017-06)
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
validate 1 other paper
1 dated earlier
1706.02677 (2017-06)
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour

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

Contrastive unsupervised learning has recently shown encouraging progress, e.g., in Momentum Contrast (MoCo) and SimCLR. In this note, we verify the effectiveness of two of SimCLR's design improvements by implementing them in the MoCo framework. With simple modifications to MoConamely, using an MLP projection head and more data augmentation-we establish stronger baselines that outperform SimCLR and do not require large training batches. We hope this will make state-of-the-art unsupervised learning research more accessible. Code will be made public.

For agents

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

get_harvested_code_for_paper("2003.04297")
get_code_for_paper("2003.04297")
have("2003.04297")

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

curl https://syntology.ai/api/ran/2003.04297.json

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

[![Syntology run record](https://syntology.ai/api/ran/2003.04297.svg)](https://syntology.ai/paper/2003.04297)

Connect an agent — have() is free.