SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2006.05479 · 2020

Probabilistic Autoencoder

arXiv · PDF · Open in the Atlas

Code that ran

7 of 13 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 5 have no contract verdict; 2 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/2006.05479.json · badge: /api/ran/2006.05479.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
vmboehm/pae-ablation canonical (classifier label)
classifier basis: census
6 ran · 6 no recorded run
chrvt/denoising-normalizing-flow extension (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
FunctionStatusWhere it lives
dequantize Ran
ran on a synthesized input
vmboehm/pae-ablation/pae/create_datasets_tf2.py
code served (permissive licence) · get_code("599a4b97af909541")
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, 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 numpy as np

# The function takes a numpy array x and adds uniform noise to it, then returns as float32
# We need to provide a numpy array as input

# Create a small numpy array as input
x = np.array([[1, 2, 3, 4],
              [5, 6, 7, 8]], dtype=np.int32)

ARGS = {
    'x': x
}
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, 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
init_once Ran
ran on a synthesized input
vmboehm/pae-ablation/pae/flow_tf2.py
code served (permissive licence) · get_code("43ef698514227548")
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 ResourceVariable. 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
import numpy as np

# Initialize TensorFlow
tf.compat.v1.disable_eager_execution()

# Create a small initializer tensor (the 'x' parameter)
x_initializer = tf.constant([[1.0, 2.0, 3.0, 4.0], [5.0, 6.0, 7.0, 8.0]])

# Define the name for the variable
variable_name = "test_variable"

ARGS = {
    "x": x_initializer,
    "name": variable_name
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ResourceVariable, shape (2, 4) — shape (2, 4) as a later run recorded it
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
integrand_circle Ran
ran; our contract draft was wrong, not the code
chrvt/denoising-normalizing-flow/experiments/evaluate.py
code served (permissive licence) · get_code("69ee46a945fa4d1a")
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 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
import torch
import torch.nn as nn

torch.manual_seed(0)

# Build a minimal model that satisfies the function's usage
class MinimalModel(nn.Module):
    def __init__(self, data_dim):
        super().__init__()
        self.data_dim = data_dim
        self.fc = nn.Linear(data_dim, data_dim)
    
    def forward(self, data, context=None, mode='dnf'):
        # The function calls: model(data.float(), context=None, mode='dnf')
        # It expects a tuple of (x_reco, log_prob_, _)
        batch_size = data.shape[0]
        x_reco = self.fc(data)
        log_prob_ = torch.randn(batch_size)
        dummy = torch.zeros(batch_size)
        return x_reco, log_prob_, dummy
    
    def eval(self):
        return super().eval()

# Create setup
model = MinimalModel(data_dim=4)
model.eval()

# z parameter: batch of 2 samples (small batch)
z = torch.linspace(0, 2 * np.pi, 2)

# data_dim parameter
data_dim = 4

# sig2 parameter
sig2 = 1.0

ARGS = {
    'z': z,
    'model': model,
    'data_dim': data_dim,
    'sig2': sig2
}
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 Tensor, 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
make_images Ran
ran on a synthesized input
vmboehm/pae-ablation/pae/model_tf2.py
code served (permissive licence) · get_code("b7596b739149d93a")
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 EagerTensor, ndim 4, 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 tensorflow as tf
import numpy as np

# Create a small batch of images
# Shape: (batch_size, width, height, depth)
batch_size = 2
width = 4
height = 3
depth = 1

# Create a tensor with shape (batch_size, width, height, depth)
images = tf.constant(
    np.random.rand(batch_size, width, height, depth).astype(np.float32)
)

nrows = 2
ncols = 1

shape = (batch_size, width, height, depth)

params = {
    'offset': False
}

ARGS = {
    'images': images,
    'nrows': nrows,
    'ncols': ncols,
    'shape': shape,
    'params': params
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type EagerTensor, ndim 4, dtype family float — no output shape is stated: observe_signature.py could not observe it (import:AttributeError)
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
residual_op Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
vmboehm/pae-ablation/AbatiComparison/models.py
code served (permissive licence) · get_code("e0fab0d1e7d95d7b")
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
import torch.nn as nn

torch.manual_seed(0)

# Create input tensor: batch_size=2, features=4
x = torch.randn(2, 4)

# Create three function modules (nn.Linear to transform features)
# f1 and f2 are required, f3 is optional
f1 = nn.Linear(4, 4)
f2 = nn.Linear(4, 4)
f3 = nn.Linear(4, 4)

functions = [f1, f2, f3]

# Create batch norm layers
# bn1 and bn2 can be present or None, bn3 should be None if f3 is None
# Using BatchNorm1d with num_features=4
bn1 = nn.BatchNorm1d(4)
bn2 = nn.BatchNorm1d(4)
bn3 = None  # satisfies: not (f3 is None and bn3 is not None)

bns = [bn1, bn2, bn3]

# Create activation function
activation_fn = nn.ReLU()

ARGS = {
    "x": x,
    "functions": functions,
    "bns": bns,
    "activation_fn": activation_fn,
}
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
spectral_norm Ran
ran on a synthesized input
vmboehm/pae-ablation/pae/util_2stageVAE_tf2.py
code served (permissive licence) · get_code("241e498a062673bc")
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 SymbolicTensor. 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
import numpy as np

# Enable eager execution for TF 1.x compatibility
tf.compat.v1.disable_eager_execution()

# Create a simple weight tensor with shape (4, 3)
# This is multi-dimensional as required by the function
weight_tensor = tf.constant(
    np.random.randn(4, 3).astype(np.float32),
    dtype=tf.float32,
    name='weight_matrix'
)

ARGS = {
    'input_': weight_tensor
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, tensorflow — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SymbolicTensor — a SymbolicTensor 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
trainable_lu_factorization Ran
ran on a synthesized input
vmboehm/pae-ablation/pae/flow_tf2.py
code served (permissive licence) · get_code("95096c3ee1dfee33")
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 tensorflow as tf

ARGS = {
    'event_size': 4,
    'batch_shape': (2,),
    'seed': 42,
    'dtype': tf.float32,
    'name': 'test_lu_factorization'
}
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 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
conv2d Not yet run
unverified (harvested, no recorded run)
vmboehm/pae-ablation/pae/util_2stageVAE_tf2.py
code served (permissive licence) · get_code("7600fd868e849534")
How it ran

no invocation recorded.

get_GN_covariance Not yet run
unverified (harvested, no recorded run)
vmboehm/pae-ablation/pae/model_tf2.py
code served (permissive licence) · get_code("a4af2f55127481cb")
How it ran

no invocation recorded.

linear Not yet run
unverified (harvested, no recorded run)
vmboehm/pae-ablation/pae/util_2stageVAE_tf2.py
code served (permissive licence) · get_code("60f7837af07944ce")
How it ran

no invocation recorded.

res_block Not yet run
unverified (harvested, no recorded run)
vmboehm/pae-ablation/pae/networks_tf2.py
code served (permissive licence) · get_code("aabec1271ea64e72")
How it ran

no invocation recorded.

res_fc_block Not yet run
unverified (harvested, no recorded run)
vmboehm/pae-ablation/pae/networks_tf2.py
code served (permissive licence) · get_code("e0d965c835d230b6")
How it ran

no invocation recorded.

scale_block Not yet run
unverified (harvested, no recorded run)
vmboehm/pae-ablation/pae/networks_tf2.py
code served (permissive licence) · get_code("0a6044204c031491")
How it ran

no invocation recorded.

Also present in other papers' code

Of 13 samples lifted from this paper's linked repositories, 13 carry a function identity and 0 do not. Of the 13 distinct functions with an identity, 2 are also present in other papers whose code we hold (1 with at least one paper dated earlier, 0 the same month, 1 later) and 11 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
integrand_circle 1 other paper
1 later
2105.12152 (2021-05)
Density estimation on low-dimensional manifolds: an inflation-deflation approach
residual_op 1 other paper
1 dated earlier
1807.01653 (2018-07)
Latent Space Autoregression for Novelty Detection

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

Principal Component Analysis (PCA) minimizes the reconstruction error given a class of linear models of fixed component dimensionality. Probabilistic PCA adds a probabilistic structure by learning the probability distribution of the PCA latent space weights, thus creating a generative model. Autoencoders (AE) minimize the reconstruction error in a class of nonlinear models of fixed latent space dimensionality and outperform PCA at fixed dimensionality. Here, we introduce the Probabilistic Autoencoder (PAE) that learns the probability distribution of the AE latent space weights using a normalizing flow (NF). The PAE is fast and easy to train and achieves small reconstruction errors, high sample quality, and good performance in downstream tasks. We compare the PAE to Variational AE (VAE), showing that the PAE trains faster, reaches a lower reconstruction error, and produces good sample quality without requiring special tuning parameters or training procedures. We further demonstrate that the PAE is a powerful model for performing the downstream tasks of probabilistic image reconstruction in the context of Bayesian inference of inverse problems for inpainting and denoising applications. Finally, we identify latent space density from NF as a promising outlier detection metric.

For agents

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

get_harvested_code_for_paper("2006.05479")
get_code_for_paper("2006.05479")
have("2006.05479")

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

curl https://syntology.ai/api/ran/2006.05479.json

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

[![Syntology run record](https://syntology.ai/api/ran/2006.05479.svg)](https://syntology.ai/paper/2006.05479)

Connect an agent — have() is free.