SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2204.02075 · 2022

Complex-Valued Autoencoders for Object Discovery

arXiv · PDF · Open in the Atlas

Code that ran

8 of 9 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 8 have no contract verdict; 0 failures of our instrument; 1 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/2204.02075.json · badge: /api/ran/2204.02075.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
loewex/complexautoencoder canonical (classifier label)
classifier basis not recorded on the link
8 ran · 1 no recorded run
FunctionStatusWhere it lives
calc_ari_score Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/utils/eval_utils.py
code served (permissive licence) · get_code("ba75ab557a965999")
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 float, 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 numpy as np
from sklearn.metrics.cluster import adjusted_rand_score

class OptConfig:
    def __init__(self):
        self.input = InputConfig()

class InputConfig:
    def __init__(self):
        self.batch_size = 2

# Create sample label arrays
# batch_size=2, spatial dimensions 4x3
batch_size = 2
height = 4
width = 3

# labels_true: shape (2, 4, 3) with values > -1 for with_background=True case
# Using values from -1 to 2 to test the filtering logic
labels_true = np.array([
    [[-1, 0, 1], [0, 1, 2], [1, 2, 1], [0, 1, 0]],
    [[0, 1, 2], [1, 2, 0], [2, 0, 1], [1, 0, 2]]
], dtype=np.int32)

# labels_pred: shape (2, 4, 3) with cluster predictions
labels_pred = np.array([
    [[0, 0, 1], [0, 1, 1], [1, 1, 0], [0, 1, 0]],
    [[0, 1, 1], [1, 1, 0], [1, 0, 1], [1, 0, 1]]
], dtype=np.int32)

opt = OptConfig()

ARGS = {
    "opt": opt,
    "labels_true": labels_true,
    "labels_pred": labels_pred,
    "with_background": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, sklearn — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, 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
get_conv_biases Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/model/model_utils.py
code served (permissive licence) · get_code("98e0779bc5e6457d")
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 torch
import torch.nn as nn
import math

torch.manual_seed(0)

ARGS = {
    "out_channels": 8,
    "fan_in": 16,
}
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 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_data Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/utils/data_utils.py
code served (permissive licence) · get_code("a47acfca52208928")
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 random
from pathlib import Path
import numpy as np
import torch
import tempfile
import os

torch.manual_seed(0)

# Create a temporary directory and npz file for testing
temp_dir = tempfile.mkdtemp()

# Create mock npz data
images_data = np.random.randn(4, 3, 2, 2).astype(np.float32)  # 4 samples, 3 channels, 2x2 spatial
labels_data = np.array([0, 1, 0, 1], dtype=np.int64)

# Save npz file
npz_path = os.path.join(temp_dir, 'test_train.npz')
np.savez(npz_path, images=images_data, labels=labels_data)

# Create a mock options object
class MockOptions:
    def __init__(self, cwd, load_path, file_name, batch_size, seed):
        self.cwd = cwd
        self.seed = seed
        self.input = MockInput(load_path, file_name, batch_size)

class MockInput:
    def __init__(self, load_path, file_name, batch_size):
        self.load_path = load_path
        self.file_name = file_name
        self.batch_size = batch_size

opt = MockOptions(
    cwd=temp_dir,
    load_path='.',
    file_name='test',
    batch_size=2,
    seed=42
)

partition = 'train'

ARGS = {
    'opt': opt,
    'partition': partition
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_dataloader Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/utils/data_utils.py
code served (permissive licence) · get_code("a499c4feff7dc1d6")
How it ran

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

import torch
import torch.utils.data
import random
import numpy as np

torch.manual_seed(0)

# Create a simple dataset
class SimpleDataset(torch.utils.data.Dataset):
    def __init__(self, size=8):
        self.size = size
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        x = torch.randn(4)
        y = torch.randint(0, 2, (1,))
        return x, y

# Create a config object with required attributes
class Config:
    def __init__(self):
        self.seed = 42
        self.input = InputConfig()

class InputConfig:
    def __init__(self):
        self.batch_size = 2

opt = Config()
dataset = SimpleDataset(size=8)

ARGS = {
    "opt": opt,
    "dataset": dataset
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_input Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/utils/data_utils.py
code served (permissive licence) · get_code("f588aa91f2172e0d")
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. 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.utils.data as data

torch.manual_seed(0)

# Create a small options object with use_cuda attribute
class Options:
    def __init__(self):
        self.use_cuda = False

opt = Options()

# Create a small dataset and data loader
class TinyDataset(data.Dataset):
    def __init__(self, num_samples=4):
        self.num_samples = num_samples
    
    def __len__(self):
        return self.num_samples
    
    def __getitem__(self, idx):
        # Return (image, label) tuple
        image = torch.randn(3, 4, 4)  # 3 channels, 4x4 spatial
        label = torch.tensor(idx % 2)
        return image, label

dataset = TinyDataset(num_samples=4)
train_loader = data.DataLoader(dataset, batch_size=2, shuffle=False)

# Create an iterator from the train_loader
iterator = iter(train_loader)

ARGS = {
    "opt": opt,
    "iterator": iterator,
    "train_loader": train_loader
}
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 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_learning_rate Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/utils/utils.py
code served (permissive licence) · get_code("606b9262e3472679")
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 float, 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

class ConfigOpt:
    def __init__(self):
        self.training = TrainingConfig()

class TrainingConfig:
    def __init__(self):
        self.learning_rate = 0.001
        self.warmup_steps = 10
        self.learning_rate_schedule = 1

opt = ConfigOpt()
step = 5

ARGS = {
    'opt': opt,
    'step': step
}
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 float, 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
get_linear_warmup_lr Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/utils/utils.py
code served (permissive licence) · get_code("9155959bd7f4b064")
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 float, 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 as nn

class OptConfig:
    def __init__(self):
        self.training = TrainingConfig()

class TrainingConfig:
    def __init__(self):
        self.warmup_steps = 100
        self.learning_rate = 0.001

opt = OptConfig()
step = 50

ARGS = {
    'opt': opt,
    'step': step
}
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 float, 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
init_channel_norm_2d Ran
ran on a synthesized input
loewex/complexautoencoder/codebase/model/model_utils.py
code served (permissive licence) · get_code("0e462099090bcc94")
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 ModuleList. 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 expects:
# - out_channel: a sequence (list) of integers representing output channels for conv layers
# - linear_out: an integer representing the output dimension for LayerNorm
# - opt: appears to be unused in the function body, but we need to provide it

out_channel = [16, 32, 64]  # List of output channel sizes for BatchNorm2d layers
linear_out = 128  # Output dimension for the final LayerNorm layer
opt = {}  # Empty dict as opt is not used in the function body

ARGS = {
    "out_channel": out_channel,
    "linear_out": linear_out,
    "opt": opt
}
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 ModuleList — a ModuleList output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a ModuleList output the recorded ndim 1 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
parse_args Not yet run
unverified (harvested, no recorded run)
loewex/complexautoencoder/codebase/utils/utils.py
code served (permissive licence) · get_code("86b5bb1b46e9cdf2")
How it ran

no invocation recorded.

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 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 9 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

Object-centric representations form the basis of human perception, and enable us to reason about the world and to systematically generalize to new settings. Currently, most works on unsupervised object discovery focus on slot-based approaches, which explicitly separate the latent representations of individual objects. While the result is easily interpretable, it usually requires the design of involved architectures. In contrast to this, we propose a comparatively simple approach - the Complex AutoEncoder (CAE) - that creates distributed object-centric representations. Following a coding scheme theorized to underlie object representations in biological neurons, its complex-valued activations represent two messages: their magnitudes express the presence of a feature, while the relative phase differences between neurons express which features should be bound together to create joint object representations. In contrast to previous approaches using complex-valued activations for object discovery, we present a fully unsupervised approach that is trained end-to-end - resulting in significant improvements in performance and efficiency. Further, we show that the CAE achieves competitive or better unsupervised object discovery performance on simple multi-object datasets compared to a state-of-the-art slot-based approach while being up to 100 times faster to train.

For agents

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

get_harvested_code_for_paper("2204.02075")
get_code_for_paper("2204.02075")
have("2204.02075")

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

curl https://syntology.ai/api/ran/2204.02075.json

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

[![Syntology run record](https://syntology.ai/api/ran/2204.02075.svg)](https://syntology.ai/paper/2204.02075)

Connect an agent — have() is free.