SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2109.13228 · NeurIPS · 2021

PASS: An ImageNet replacement for self-supervised pretraining without humans

Andrea Vedaldi, Andrew Zisserman, Yuki Asano, Christian Rupprecht

arXiv · PDF · Open in the Atlas

Code that ran

11 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 3 honoured the contract we drafted, 2 violated it and 0 have no contract verdict; 6 failures of our instrument; 3 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/2109.13228.json · badge: /api/ran/2109.13228.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
facebookresearch/dino canonical (classifier label)
classifier basis not recorded on the link
11 ran · 3 no recorded run
FunctionStatusWhere it lives
apply_mask Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
facebookresearch/dino/visualize_attention.py
code served (permissive licence) · get_code("c83fb22106327e13")
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 float, shape (4, 3, 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 torch
import numpy as np

torch.manual_seed(0)

# Create a small image array (height=4, width=3, channels=3)
image = np.random.rand(4, 3, 3).astype(np.float32) * 255

# Create a small mask array (height=4, width=3) with values between 0 and 1
mask = np.random.rand(4, 3).astype(np.float32)

# Create a color array with 3 channel values (RGB)
color = np.array([1.0, 0.5, 0.2])

# alpha parameter for blending
alpha = 0.5

ARGS = {
    'image': image,
    'mask': mask,
    'color': color,
    'alpha': alpha
}
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 ndarray, ndim 3, dtype family float, shape (4, 3, 3) — shape (4, 3, 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
bool_flag Ran
ran, violated the contract we drafted
facebookresearch/dino/utils.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from sthalles/MaSSL, is pointer only (licence: NONE) · get_code("c19231378e41dc5a")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from sthalles/MaSSL, 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.

clip_gradients Ran
ran, honoured the contract we drafted
facebookresearch/dino/utils.py
code served (permissive licence) · get_code("2b5489053b222e00")
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
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
    
    def forward(self, x):
        return self.linear2(self.linear1(x))

model = SimpleModel()

# Create a simple forward pass and backward pass to populate gradients
x = torch.randn(2, 4)
y = torch.randn(2, 2)
output = model(x)
loss = nn.MSELoss()(output, y)
loss.backward()

clip = 1.0

ARGS = {
    "model": model,
    "clip": clip,
}
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
config_imname Ran
ran; our contract draft was wrong, not the code
facebookresearch/dino/eval_image_retrieval.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("49cf1f4ce52e9613")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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.

config_qimname Ran
ran; our contract draft was wrong, not the code
facebookresearch/dino/eval_image_retrieval.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("91c5ae2396603107")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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.

cosine_scheduler Ran
ran, honoured the contract we drafted
facebookresearch/dino/utils.py
code served (permissive licence) · get_code("361a6b24f11fc50a")
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 ndarray, ndim 1, dtype family float, shape (500,) 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

ARGS = {
    "base_value": 0.1,
    "final_value": 0.001,
    "epochs": 10,
    "niter_per_ep": 50,
    "warmup_epochs": 2,
    "start_warmup_value": 0.0,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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 (500,) — shape (500,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
drop_path Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
facebookresearch/dino/vision_transformer.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from gladzhang/Xformer, is pointer only (licence: NONE) · get_code("55120f2026b56aa2")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from gladzhang/Xformer, 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.

is_image_file Ran
ran, violated the contract we drafted
facebookresearch/dino/eval_copy_detection.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("dfebfd56865b6379")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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.

norm_mask Ran
ran; our contract draft was wrong, not the code
facebookresearch/dino/eval_video_segmentation.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("e015f2c93c762aff")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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.

random_colors Ran
ran, honoured the contract we drafted
facebookresearch/dino/visualize_attention.py
code served (permissive licence) · get_code("cc71140102574a2c")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 2, dtype family 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 random
import colorsys

random.seed(0)

# Construct ARGS for random_colors function
ARGS = {
    "N": 5,
    "bright": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, 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
score_ap_from_ranks_1 Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
facebookresearch/dino/eval_copy_detection.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("5f21619ff716c1ce")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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.

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

no invocation recorded.

label_propagation Not yet run
unverified (harvested, no recorded run)
facebookresearch/dino/eval_video_segmentation.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("3093e0466654d790")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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. No invocation recorded.

restrict_neighborhood Not yet run
unverified (harvested, no recorded run)
facebookresearch/dino/eval_video_segmentation.py
pointer only (licence: Apache-2.0); the stored copy get_code serves, from mahmoodlab/HIPT, is pointer only (licence: NOASSERTION) · get_code("efc78e753491eeab")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (Apache-2.0), but get_code serves the stored copy, first harvested from mahmoodlab/HIPT, 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. No invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 14 are also present in other papers whose code we hold (13 with at least one paper dated earlier, 1 the same month, 14 later) and 0 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
drop_path 224 other papers
12 dated earlier, 1 the same month, 210 later, 1 undated · 38 only as an assembly brick
2003.12039 (2020-03)
RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
2007.09384 (2020-07)
Multi-Scale Positive Sample Refinement for Few-Shot Object Detection
2010.11929 (2020-10)
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
and 221 more not named here
cosine_scheduler 37 other papers
3 dated earlier, 34 later
2005.10242 (2020-05)
Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere
2103.15375 (2021-03)
AlignMixup: Improving Representations By Interpolating Aligned Features
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
and 34 more not named here
bool_flag 31 other papers
1 dated earlier, 30 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2111.12958 (2021-11)
Self-Distilled Self-Supervised Representation Learning
136810551 (2022 (year only))
UniMiSS: Universal Medical Self-Supervised Learning via Breaking Dimensionality Barrier
and 28 more not named here
clip_gradients 28 other papers
1 dated earlier, 27 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2111.12958 (2021-11)
Self-Distilled Self-Supervised Representation Learning
136810551 (2022 (year only))
UniMiSS: Universal Medical Self-Supervised Learning via Breaking Dimensionality Barrier
and 25 more not named here
random_colors 9 other papers
5 dated earlier, 4 later
1612.03242 (2016-12)
StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks
1702.01983 (2017-02)
Face Aging With Conditional Generative Adversarial Networks
and 6 more not named here
config_imname 4 other papers
2 dated earlier, 2 later
1711.02512 (2017-11)
Fine-tuning CNN Image Retrieval with No Human Annotation
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2308.04008 (2023-08)
Coarse-to-Fine: Learning Compact Discriminative Representation for Single-Stage Image Retrieval
and 1 more not named here
config_qimname 4 other papers
2 dated earlier, 2 later
1711.02512 (2017-11)
Fine-tuning CNN Image Retrieval with No Human Annotation
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2308.04008 (2023-08)
Coarse-to-Fine: Learning Compact Discriminative Representation for Single-Stage Image Retrieval
and 1 more not named here
knn_classifier 4 other papers
2 dated earlier, 2 later
2104.14294 (2021-04)
Emerging Properties in Self-Supervised Vision Transformers
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2410.05057 (2024-10)
SELECT: A Large-Scale Benchmark of Data Curation Strategies for Image Classification
and 1 more not named here
norm_mask 3 other papers
1 dated earlier, 2 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2306.03881 (2023-06)
Emergent Correspondence from Image Diffusion
2602.01951 (2026-02)
Enabling Progressive Whole-slide Image Analysis with Multi-scale Pyramidal Network
restrict_neighborhood 3 other papers
1 dated earlier, 2 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2306.03881 (2023-06)
Emergent Correspondence from Image Diffusion
2602.01951 (2026-02)
Enabling Progressive Whole-slide Image Analysis with Multi-scale Pyramidal Network
score_ap_from_ranks_1 3 other papers
1 dated earlier, 2 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2202.10261 (2022-02)
A Self-Supervised Descriptor for Image Copy Detection
2602.01951 (2026-02)
Enabling Progressive Whole-slide Image Analysis with Multi-scale Pyramidal Network
is_image_file 2 other papers
1 dated earlier, 1 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2602.01951 (2026-02)
Enabling Progressive Whole-slide Image Analysis with Multi-scale Pyramidal Network
label_propagation 2 other papers
1 dated earlier, 1 later
2106.09681 (2021-06)
XCiT: Cross-Covariance Image Transformers
2602.01951 (2026-02)
Enabling Progressive Whole-slide Image Analysis with Multi-scale Pyramidal Network
apply_mask 1 other paper
1 later
2311.15264 (2023-11)
ChAda-ViT : Channel Adaptive Attention for Joint Representation Learning of Heterogeneous Microscopy Images

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

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

Abstract

Computer vision has long relied on ImageNet and other large datasets of images sampled from the Internet for pretraining models. However, these datasets have ethical and technical shortcomings, such as containing personal information taken without consent, unclear license usage, biases, and, in some cases, even problematic image content. On the other hand, state-of-the-art pretraining is nowadays obtained with unsupervised methods, meaning that labelled datasets such as ImageNet may not be necessary, or perhaps not even optimal, for model pretraining. We thus propose an unlabelled dataset PASS: Pictures without humAns for Self-Supervision. PASS only contains images with CC-BY license and complete attribution metadata, addressing the copyright issue. Most importantly, it contains no images of people at all, and also avoids other types of images that are problematic for data protection or ethics. We show that PASS can be used for pretraining with methods such as MoCo-v2, SwAV and DINO. In the transfer learning setting, it yields similar downstream performances to ImageNet pretraining even on tasks that involve humans, such as human pose estimation. PASS does not make existing datasets obsolete, as for instance it is insufficient for benchmarking. However, it shows that model pretraining is often possible while using safer data, and it also provides the basis for a more robust evaluation of pretraining methods. Dataset and pretrained models are available 1 .

For agents

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

get_harvested_code_for_paper("2109.13228")
get_code_for_paper("2109.13228")
have("2109.13228")

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

curl https://syntology.ai/api/ran/2109.13228.json

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

[![Syntology run record](https://syntology.ai/api/ran/2109.13228.svg)](https://syntology.ai/paper/2109.13228)

Connect an agent — have() is free.