SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2104.14294 · ICCV · 2021

Emerging Properties in Self-Supervised Vision Transformers

Mathilde Caron, Ishan Misra, Julien Mairal, Piotr Bojanowski, Armand Joulin, Hugo Touvron, Hervé Jegou

arXiv · PDF · Open in the Atlas

Code that ran

14 of 20 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 9 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/2104.14294.json · badge: /api/ran/2104.14294.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
sithu31296/simple-object-tracking community (classifier label)
classifier basis not recorded on the link
6 ran · 3 no recorded run
ipmi-icns-uke/sparsam community (classifier label)
classifier basis: plan_row
5 ran · 1 no recorded run
jankrepl/mildlyoverfitted reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
clemsgrs/hipt reimplementation (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
copy not recorded community
not in the paper's declared repositories
1 ran · 0 no recorded run
alijavidani/local_global_representation_learning extension (classifier label)
classifier basis: classifier_2025
0 ran · 1 no recorded run
FunctionStatusWhere it lives
checkpoint_filter_fn Ran
ran; our contract draft was wrong, not the code
ipmi-icns-uke/sparsam/sparsam/models.py
pointer only (licence: MIT); the stored copy get_code serves, from NVlabs/FAN, is pointer only (licence: NOASSERTION) · get_code("e25a3f7ad14df60b")
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 NVlabs/FAN, and GitHub could not classify the licence of that repository (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.

coco_class_index Ran
ran on a synthesized input
sithu31296/simple-object-tracking/tracking/utils.py
code served (permissive licence) · get_code("1143ac599d89fc39")
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.

# Setup code for coco_class_index function

ARGS = {
    'class_name': 'person'
}
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
compute_embedding Ran
ran; our contract draft was wrong, not the code
jankrepl/mildlyoverfitted/github_adventures/dino/evaluation.py
code served (permissive licence) · get_code("469756a4afefa152")
How it ran

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

torch.manual_seed(0)

# Create a minimal Vision Transformer backbone
class MinimalViT(nn.Module):
    def __init__(self, embed_dim=4):
        super().__init__()
        self.embed_dim = embed_dim
        # Simple linear layer to project to embedding dimension
        self.proj = nn.Linear(3 * 8 * 8, embed_dim)
    
    def forward(self, x):
        # x shape: (batch, 3, 8, 8)
        batch_size = x.shape[0]
        x = x.view(batch_size, -1)  # Flatten
        x = self.proj(x)  # Project to embed_dim
        return x

backbone = MinimalViT(embed_dim=4)
backbone.eval()

# Create a minimal dataset class
class MinimalDataset:
    def __init__(self, num_samples=2, img_height=8, img_width=8):
        self.num_samples = num_samples
        self.img_height = img_height
        self.img_width = img_width
        self.classes = ["class_0", "class_1"]
    
    def __len__(self):
        return self.num_samples
    
    def __getitem__(self, idx):
        # Return (image, label)
        img = torch.randn(3, self.img_height, self.img_width)
        label = idx % 2
        return img, label

# Create a minimal data loader
dataset = MinimalDataset(num_samples=2, img_height=8, img_width=8)

# Manually create a data loader as a list of batches
data_list = []
for i in range(2):
    img, label = dataset[i]
    data_list.append((img.unsqueeze(0), torch.tensor([label])))

# Combine into batches
batch_imgs = torch.cat([item[0] for item in data_list], dim=0)
batch_labels = torch.cat([item[1] for item in data_list], dim=0)

class SimpleDataLoader:
    def __init__(self, images, labels, dataset):
        self.images = images
        self.labels = labels
        self.dataset = dataset
    
    def __iter__(self):
        yield (self.images, self.labels)

data_loader = SimpleDataLoader(batch_imgs, batch_labels, dataset)

ARGS = {
    "backbone": backbone,
    "data_loader": data_loader,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple 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
filter_classes Ran
ran on a synthesized input
ipmi-icns-uke/sparsam/sparsam/helper.py
code served (permissive licence) · get_code("adcec324a2226ec9")
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. 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

# Create sample features and labels
features = np.array([
    [1.0, 2.0, 3.0, 4.0],
    [5.0, 6.0, 7.0, 8.0],
    [9.0, 10.0, 11.0, 12.0],
    [13.0, 14.0, 15.0, 16.0],
    [17.0, 18.0, 19.0, 20.0],
    [21.0, 22.0, 23.0, 24.0]
])

labels = np.array([0, 1, 0, 2, 1, 2])

filter_labels = np.array([0, 1])

ARGS = {
    "features": features,
    "labels": labels,
    "filter_labels": filter_labels
}
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 tuple — a tuple 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
get_args_parser Ran
ran; our contract draft was wrong, not the code
clemsgrs/hipt/src/train/classification.py
pointer only (licence: NONE) · get_code("6dfcca5f4b2d0098")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, clemsgrs/hipt 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_large_classes Ran
ran on a synthesized input
ipmi-icns-uke/sparsam/sparsam/helper.py
code served (permissive licence) · get_code("daa26eb67407e2ea")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 1, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

ARGS = {
    "labels": np.array([0, 1, 1, 2, 2, 2, 3, 3, 3, 3]),
    "min_amount": 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 list, ndim 1, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_params_groups Ran
ran; our contract draft was wrong, not the code
ipmi-icns-uke/sparsam/sparsam/helper.py
code served (permissive licence) · get_code("ec22c16f1e0653f7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned 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
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        # Create parameters with different shapes to test the logic
        self.linear1 = nn.Linear(4, 3)  # weight: (3, 4), bias: (3,)
        self.linear2 = nn.Linear(3, 2)  # weight: (2, 3), bias: (2,)
        self.register_parameter('scalar_param', nn.Parameter(torch.tensor(1.0)))
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

ARGS = {
    'model': model
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type 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
iou Ran
ran on a synthesized input
sithu31296/simple-object-tracking/tracking/sort/matching.py
code served (permissive licence) · get_code("4089a9837aebe640")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family float, shape (2,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# bbox is a 1D array with format (top_left_x, top_left_y, width, height)
bbox = np.array([10.0, 20.0, 30.0, 40.0], dtype=np.float32)

# candidates is a 2D array where each row is (top_left_x, top_left_y, width, height)
candidates = np.array([
    [15.0, 25.0, 20.0, 30.0],
    [5.0, 10.0, 50.0, 60.0]
], dtype=np.float32)

ARGS = {
    "bbox": bbox,
    "candidates": candidates
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family float, shape (2,) — shape (2,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
iou_cost Ran
ran on a synthesized input
sithu31296/simple-object-tracking/tracking/sort/matching.py
code served (permissive licence) · get_code("b71fec04b665a291")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 2, dtype family float, shape (2, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

class Track:
    def __init__(self, tlwh, time_since_update=0):
        self.tlwh_value = tlwh
        self.time_since_update = time_since_update
    
    def to_tlwh(self):
        return self.tlwh_value

class Detection:
    def __init__(self, tlwh):
        self.tlwh = tlwh

# Create sample tracks
track1 = Track(np.array([10.0, 20.0, 50.0, 60.0]), time_since_update=0)
track2 = Track(np.array([100.0, 120.0, 40.0, 50.0]), time_since_update=2)

# Create sample detections
detection1 = Detection(np.array([15.0, 25.0, 48.0, 58.0]))
detection2 = Detection(np.array([105.0, 125.0, 42.0, 52.0]))
detection3 = Detection(np.array([200.0, 200.0, 30.0, 40.0]))

tracks = [track1, track2]
detections = [detection1, detection2, detection3]

ARGS = {
    "tracks": tracks,
    "detections": detections,
    "track_indices": None,
    "detection_indices": None
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 2, dtype family float, shape (2, 3) — shape (2, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
letterbox Ran
ran on a synthesized input
sithu31296/simple-object-tracking/tracking/utils.py
code served (permissive licence) · get_code("128c6d6cc48dcf20")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 3, dtype family int, shape (480, 640, 3) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import cv2
import numpy as np

# Create a small test image (BGR format as expected by cv2)
# Using a 480x640 image to have distinct dimensions
img = np.random.randint(0, 255, (480, 640, 3), dtype=np.uint8)

# Define ARGS dict with all parameters
ARGS = {
    "img": img,
    "new_shape": (640, 640)
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
cv2, numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 3, dtype family int, shape (480, 640, 3) — shape (480, 640, 3) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
min_cost_matching Ran
ran on a synthesized input
sithu31296/simple-object-tracking/tracking/sort/matching.py
code served (permissive licence) · get_code("6278dd3b36b3eadc")
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. 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

# Define minimal stand-in classes for Track and Detection
class Track:
    def __init__(self, track_id, state):
        self.track_id = track_id
        self.state = state

class Detection:
    def __init__(self, detection_id, feature):
        self.detection_id = detection_id
        self.feature = feature

# Define a distance metric function that computes cost matrix
def distance_metric(tracks, detections, track_indices, detection_indices):
    """
    Compute NxM cost matrix where N = len(track_indices), M = len(detection_indices).
    Returns a matrix with distinct dimensions.
    """
    n_tracks = len(track_indices)
    n_detections = len(detection_indices)
    # Create a simple cost matrix with values between 0 and 10
    cost_matrix = np.random.rand(n_tracks, n_detections) * 8.0
    return cost_matrix

# Create sample tracks and detections
tracks = [
    Track(0, np.array([1.0, 2.0, 3.0, 4.0])),
    Track(1, np.array([5.0, 6.0, 7.0, 8.0])),
    Track(2, np.array([9.0, 10.0, 11.0, 12.0])),
]

detections = [
    Detection(0, np.array([1.1, 2.1, 3.1, 4.1])),
    Detection(1, np.array([5.1, 6.1, 7.1, 8.1])),
]

# Define arguments
ARGS = {
    "distance_metric": distance_metric,
    "max_distance": 5.0,
    "tracks": tracks,
    "detections": detections,
    "track_indices": None,
    "detection_indices": None,
}
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 tuple — a tuple 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
min_max_normalize_tensor Ran
ran on a synthesized input
ipmi-icns-uke/sparsam/sparsam/utils.py
code served (permissive licence) · get_code("9d228d3aaa15dc45")
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)

# Create a small tensor with values that will be clamped and normalized
img = torch.randn(2, 4) * 10  # Random values scaled to be larger

# Define normalization parameters
min_value = -5.0
max_value = 5.0

ARGS = {
    "img": img,
    "min_value": min_value,
    "max_value": max_value,
}
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
pair Ran
ran, violated the contract we drafted
this paper's copy was not recorded; identical code first harvested from LINs-lab/DynMoE
pointer only · get_code("6ba8cee9f5daea41")
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.

scale_boxes Ran
ran on a synthesized input
sithu31296/simple-object-tracking/tracking/utils.py
code served (permissive licence) · get_code("55e839c9dc92868b")
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)

# Create a small batch of bounding boxes
# boxes shape: (batch_size, num_boxes_info)
# The function uses boxes[:, ::2], boxes[:, 1::2], boxes[:, :4]
# So we need at least 4 columns for boxes
boxes = torch.tensor([
    [10.0, 20.0, 50.0, 60.0],
    [15.0, 25.0, 55.0, 65.0]
], dtype=torch.float32)

# Original image shape (Height, Width)
orig_shape = (100, 120)

# New shape after scaling (Height, Width)
new_shape = (200, 240)

ARGS = {
    "boxes": boxes,
    "orig_shape": orig_shape,
    "new_shape": new_shape
}
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
build_model Not yet run
unverified (harvested, no recorded run)
sithu31296/simple-object-tracking/tracking/clip/model.py
pointer only (licence: MIT); the stored copy get_code serves, from tmlr-group/CARPRT, is pointer only (licence: NONE) · get_code("f3830274a7e18d5e")
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 tmlr-group/CARPRT, 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.

compute_knn Not yet run
unverified (harvested, no recorded run)
jankrepl/mildlyoverfitted/github_adventures/dino/evaluation.py
code served (permissive licence) · get_code("27951b936651c6e9")
How it ran

no invocation recorded.

knn_classifier Not yet run
unverified (harvested, no recorded run)
alijavidani/local_global_representation_learning/eval_knn.py
code served (permissive licence) · get_code("6ecb6791e6a9ebfb")
How it ran

no invocation recorded.

load Not yet run
unverified (harvested, no recorded run)
sithu31296/simple-object-tracking/tracking/clip/clip.py
code served (permissive licence) · get_code("acaee8a26cc4ebcc")
How it ran

no invocation recorded.

load Not yet run
unverified (harvested, no recorded run)
sithu31296/simple-object-tracking/tracking/dino/dino.py
code served (permissive licence) · get_code("f0765c91f6e94bc8")
How it ran

no invocation recorded.

model_inference Not yet run
unverified (harvested, no recorded run)
ipmi-icns-uke/sparsam/sparsam/utils.py
code served (permissive licence) · get_code("3c5097faa5cd9b8b")
How it ran

no invocation recorded.

Also present in other papers' code

Of 20 samples lifted from this paper's linked repositories, 20 carry a function identity and 0 do not. Of the 20 distinct functions with an identity, 8 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 8 later) and 12 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
build_model 119 other papers
2 dated earlier, 115 later, 2 undated
1809.11096 (2018-09)
Large Scale GAN Training for High Fidelity Natural Image Synthesis
1901.07441 (2019-01)
PadChest: A large chest x-ray image dataset with multi-label annotated reports
2107.06383 (2021-07)
How Much Can CLIP Benefit Vision-and-Language Tasks?
and 116 more not named here
pair 90 other papers
4 dated earlier, 85 later, 1 undated · 8 only as an assembly brick
2004.13649 (2020-04)
Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels
2010.11929 (2020-10)
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
2103.15691 (2021-03)
ViViT: A Video Vision Transformer
and 87 more not named here
get_params_groups 10 other papers
10 later · 1 only as an assembly brick
2211.11727 (2022-11)
Parametric Classification for Generalized Category Discovery: A Baseline Study
2308.12112 (2023-08)
Category Adaptation Meets Projected Distillation in Generalized Continual Category Discovery
2310.01376 (2023-10)
Towards Distribution-Agnostic Generalized Category Discovery
and 7 more not named here
knn_classifier 4 other papers
4 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2109.13228 (2021-09)
PASS: An ImageNet replacement for self-supervised pretraining without humans
2410.05057 (2024-10)
SELECT: A Large-Scale Benchmark of Data Curation Strategies for Image Classification
and 1 more not named here
checkpoint_filter_fn 1 other paper
1 later
2410.02675 (2024-10)
FAN: Fourier Analysis Networks
letterbox 1 other paper
1 later
2107.11291 (2021-07)
Human Pose Regression with Residual Log-likelihood Estimation
min_cost_matching 1 other paper
1 later
2202.13514 (2022-02)
StrongSORT: Make DeepSORT Great Again
scale_boxes 1 other paper
1 later
2107.11291 (2021-07)
Human Pose Regression with Residual Log-likelihood Estimation

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

Figure 1: Self-attention from a Vision Transformer with 8 × 8 patches trained with no supervision. We look at the self-attention of the [CLS] token on the heads of the last layer. This token is not attached to any label nor supervision. These maps show that the model automatically learns class-specific features leading to unsupervised object segmentations.

For agents

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

get_harvested_code_for_paper("2104.14294")
get_code_for_paper("2104.14294")
have("2104.14294")

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

curl https://syntology.ai/api/ran/2104.14294.json

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

[![Syntology run record](https://syntology.ai/api/ran/2104.14294.svg)](https://syntology.ai/paper/2104.14294)

Connect an agent — have() is free.