SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.17896 · 2023

Exploring the Limits of Deep Image Clustering using Pretrained Models

arXiv · PDF · Open in the Atlas

Code that ran

4 of 4 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 4 failures of our instrument; 0 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/2303.17896.json · badge: /api/ran/2303.17896.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
HHU-MMBS/TEMI-official-BMVC2023 canonical (classifier label)
classifier basis: classifier_2025
4 ran · 0 no recorded run
FunctionStatusWhere it lives
compute_embedding Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
HHU-MMBS/TEMI-official-BMVC2023/gen_embeds.py
code served (permissive licence) · get_code("a4e72aa97d72ef5e")
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. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type 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 model that takes images and returns embeddings
class SimpleEmbeddingModel(nn.Module):
    def __init__(self, input_channels=3, embedding_dim=4):
        super().__init__()
        self.fc = nn.Linear(input_channels * 8 * 8, embedding_dim)
    
    def forward(self, x):
        # x shape: (batch, 3, 8, 8)
        x = x.view(x.size(0), -1)  # flatten to (batch, 192)
        return self.fc(x)  # returns (batch, 4)

model = SimpleEmbeddingModel(input_channels=3, embedding_dim=4)
model.eval()

# Create a minimal data loader as a list of (images, labels) tuples
# Each image is (3, 8, 8), labels are scalars or 1D tensors
loader = [
    (torch.randn(2, 3, 8, 8), torch.tensor([0, 1])),
    (torch.randn(2, 3, 8, 8), torch.tensor([1, 0])),
]

ARGS = {
    "model": model,
    "loader": loader,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch, tqdm — 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
compute_neighbors Ran
ran; our contract draft was wrong, not the code
HHU-MMBS/TEMI-official-BMVC2023/gen_embeds.py
code served (permissive licence) · get_code("709a9a2f23ee145d")
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 topk. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type 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 embedding tensor
# Shape: (num_embeddings, embedding_dim)
# Using small dimensions: 6 embeddings of dimension 4
embedding = torch.randn(6, 4)

# k parameter for top-k neighbors
k = 2

ARGS = {
    "embedding": embedding,
    "k": k,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type topk — a topk output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a topk output the recorded ndim 3 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
default_out_dir Ran
ran; our contract draft was wrong, not the code
HHU-MMBS/TEMI-official-BMVC2023/train_main.py
code served (permissive licence) · get_code("79b303c122ae61f6")
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 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 torch
import torch.nn as nn

torch.manual_seed(0)

# The function default_out_dir takes two string parameters: loss and dset
# It simply formats them into a path string, so we need valid string values

ARGS = {
    'loss': 'mse',
    'dset': 'cifar10'
}
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 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_outpath Ran
ran; our contract draft was wrong, not the code
HHU-MMBS/TEMI-official-BMVC2023/gen_embeds.py
code served (permissive licence) · get_code("5914e91451bf1c5e")
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 PosixPath. 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
from pathlib import Path

# Analyze the function:
# - arch: string, used with .replace('/', '_')
# - dataset: string, used with .replace('/', '_')
# - datapath: string (default 'data'), used with Path().expanduser().resolve()
# All parameters are strings, no complex objects needed

ARGS = {
    'arch': 'resnet50',
    'dataset': 'imagenet',
    'datapath': 'data'
}
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 PosixPath — a PosixPath output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a PosixPath 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

Also present in other papers' code

Of 4 samples lifted from this paper's linked repositories, 4 carry a function identity and 0 do not. Of the 4 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 4 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.

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

We present a general methodology that learns to classify images without labels by leveraging pretrained feature extractors. Our approach involves self-distillation training of clustering heads based on the fact that nearest neighbours in the pretrained feature space are likely to share the same label. We propose a novel objective that learns associations between image features by introducing a variant of pointwise mutual information together with instance weighting. We demonstrate that the proposed objective is able to attenuate the effect of false positive pairs while efficiently exploiting the structure in the pretrained feature space. As a result, we improve the clustering accuracy over $k$-means on $17$ different pretrained models by $6.1$\% and $12.2$\% on ImageNet and CIFAR100, respectively. Finally, using self-supervised vision transformers, we achieve a clustering accuracy of $61.6$\% on ImageNet. The code is available at https://github.com/HHU-MMBS/TEMI-official-BMVC2023.

For agents

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

get_harvested_code_for_paper("2303.17896")
get_code_for_paper("2303.17896")
have("2303.17896")

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

curl https://syntology.ai/api/ran/2303.17896.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.17896.svg)](https://syntology.ai/paper/2303.17896)

Connect an agent — have() is free.