SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1807.07543 · 2018

Understanding and Improving Interpolation in Autoencoders via an Adversarial Regularizer

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; 1 honoured the contract we drafted, 1 violated it and 3 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/1807.07543.json · badge: /api/ran/1807.07543.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
anonymous-iclr-2019/acai-iclr-2019 canonical (classifier label)
classifier basis: census
2 ran · 2 no recorded run
ivanvovk/controllable-face-generation community (classifier label)
classifier basis not recorded on the link
7 ran · 1 no recorded run
copy not recorded community
not in the paper's declared repositories
2 ran · 0 no recorded run
FunctionStatusWhere it lives
avg_losses Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from baohq1595/aae-experiment
pointer only; the stored copy get_code serves, from baohq1595/aae-experiment, is pointer only (licence: NONE) · get_code("3d2866ba4046ca98")
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.

draw_line Ran
ran on a synthesized input
anonymous-iclr-2019/acai-iclr-2019/lib/data.py
code served (permissive licence) · get_code("d1676a94e8a860db")
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 (32, 32, 1) 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 math
import numpy as np

ARGS = {
    "angle": 0.785,  # approximately 45 degrees in radians
    "height": 32,
    "width": 32,
    "w": 2.0
}
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 3, dtype family float, shape (32, 32, 1) — shape (32, 32, 1) 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
input_lines Ran
ran on a synthesized input
anonymous-iclr-2019/acai-iclr-2019/lib/data.py
code served (permissive licence) · get_code("d920996d38490ef5")
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 math
import random
import numpy as np
import tensorflow as tf

ARGS = {
    "batch_size": 2,
    "size": (4, 3, 1),
    "limit": 4
}
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 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
is_sequence Ran
ran, violated the contract we drafted
ivanvovk/controllable-face-generation/lreq.py
pointer only (licence: MIT); the stored copy get_code serves, from podgorskiy/StyleGAN_Blobless, is pointer only (licence: NONE) · get_code("eb766ff11551476d")
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 podgorskiy/StyleGAN_Blobless, 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.

kl Ran
ran; our contract draft was wrong, not the code
ivanvovk/controllable-face-generation/losses.py
code served (permissive licence) · get_code("96e5a2bdc93b101d")
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 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

torch.manual_seed(0)

# Create small tensors for mu and log_var
# The function performs operations: pow(2), exp(), mean
# Shapes should be (batch_size, latent_dim) for a VAE-like usage
batch_size = 2
latent_dim = 4

mu = torch.randn(batch_size, latent_dim)
log_var = torch.randn(batch_size, latent_dim)

ARGS = {
    "mu": mu,
    "log_var": log_var,
}
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 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
make_image_grid Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
this paper's copy was not recorded; identical code first harvested from baohq1595/aae-experiment
pointer only; the stored copy get_code serves, from baohq1595/aae-experiment, is pointer only (licence: NONE) · get_code("73a2b3dfb83b794b")
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.

make_tuple Ran
ran, honoured the contract we drafted
ivanvovk/controllable-face-generation/lreq.py
pointer only (licence: MIT); the stored copy get_code serves, from podgorskiy/StyleGAN_Blobless, is pointer only (licence: NONE) · get_code("509d9bfd9a1c5420")
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 podgorskiy/StyleGAN_Blobless, 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.

pixel_norm Ran
ran; our contract draft was wrong, not the code
ivanvovk/controllable-face-generation/modules.py
code served (permissive licence) · get_code("a320e6151da0b774")
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)

# The function takes a tensor x and an optional epsilon parameter
# It performs pixel normalization: normalizes each pixel across the channel dimension
# x should be a tensor where mean is computed over dim=1

x = torch.randn(2, 4)  # batch_size=2, channels=4
epsilon = 1e-08

ARGS = {
    "x": x,
    "epsilon": epsilon,
}
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
reconstruction Ran
ran; our contract draft was wrong, not the code
ivanvovk/controllable-face-generation/losses.py
code served (permissive licence) · get_code("d1629b5dc1a4aac0")
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 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

torch.manual_seed(0)

# Create small tensors for reconstruction input and target
batch_size = 2
dim = 4

recon_x = torch.randn(batch_size, dim)
x = torch.randn(batch_size, dim)
lod = None

ARGS = {
    "recon_x": recon_x,
    "x": x,
    "lod": lod,
}
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 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
style_mod Ran
ran on a synthesized input
ivanvovk/controllable-face-generation/modules.py
code served (permissive licence) · get_code("01cc5106835be4f8")
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 4, dtype family float, shape (2, 4, 3, 5) 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)

# Analyze the function:
# - x: tensor with shape (batch, channels, height, width) - used with .device and indexed
# - style: tensor that needs:
#   - .to(x.device) method
#   - .view() method
#   - shape[0] (batch dimension)
#   - indexing with [:, 1] and [:, 0]
#   - Final shape should be (batch, 2, x.shape[1], 1, 1)
#
# The function uses torch.addcmul which takes:
#   - input tensor: style[:, 1]
#   - tensor1: x
#   - tensor2: style[:, 0] + 1
#   - value: 1.0
#
# So x and style[:, 0] must have compatible shapes for multiplication
# style[:, 1] must have compatible shape for addition

batch_size = 2
channels = 4

# x should be (batch, channels, height, width)
x = torch.randn(batch_size, channels, 3, 5)

# style needs to reshape to (batch, 2, channels, 1, 1)
# Before reshaping, it should have batch_size * 2 * channels elements
# We can create it as (batch, 2*channels) and it will view to (batch, 2, channels, 1, 1)
style = torch.randn(batch_size, 2 * channels)

ARGS = {
    "x": x,
    "style": style,
}
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 4, dtype family float, shape (2, 4, 3, 5) — shape (2, 4, 3, 5) 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
upscale2d Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
ivanvovk/controllable-face-generation/modules.py
pointer only (licence: MIT); the stored copy get_code serves, from podgorskiy/StyleGAN_Blobless, is pointer only (licence: NONE) · get_code("10f278fd8f76579c")
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 podgorskiy/StyleGAN_Blobless, 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.

discriminator_logistic_simple_gp Not yet run
unverified (harvested, no recorded run)
ivanvovk/controllable-face-generation/losses.py
code served (permissive licence) · get_code("2d1d3999674791df")
How it ran

no invocation recorded.

get_latents_and_labels Not yet run
unverified (harvested, no recorded run)
anonymous-iclr-2019/acai-iclr-2019/cluster.py
code served (permissive licence) · get_code("9d0f581182913f69")
How it ran

no invocation recorded.

single_layer_classifier Not yet run
unverified (harvested, no recorded run)
anonymous-iclr-2019/acai-iclr-2019/lib/classifiers.py
code served (permissive licence) · get_code("0f185dd0ad0c3ed4")
How it ran

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, 8 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 6 later) and 6 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
upscale2d 6 other papers
6 later
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
2310.01174 (2023-10)
Light Schrödinger Bridge
2402.03207 (2024-02)
Light and Optimal Schrödinger Bridge Matching
and 3 more not named here
kl 5 other papers
5 later
2310.01174 (2023-10)
Light Schrödinger Bridge
2402.03207 (2024-02)
Light and Optimal Schrödinger Bridge Matching
2511.06856 (2025-11)
Contact Wasserstein Geodesics for Non-Conservative Schrödinger Bridges
and 2 more not named here
pixel_norm 5 other papers
5 later
2310.01174 (2023-10)
Light Schrödinger Bridge
2402.03207 (2024-02)
Light and Optimal Schrödinger Bridge Matching
2511.06856 (2025-11)
Contact Wasserstein Geodesics for Non-Conservative Schrödinger Bridges
and 2 more not named here
reconstruction 5 other papers
5 later
2310.01174 (2023-10)
Light Schrödinger Bridge
2402.03207 (2024-02)
Light and Optimal Schrödinger Bridge Matching
2511.06856 (2025-11)
Contact Wasserstein Geodesics for Non-Conservative Schrödinger Bridges
and 2 more not named here
is_sequence 2 other papers
2 later
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
2605.11983 (2026-05)
QDSB: Quantized Diffusion Schrödinger Bridges
make_tuple 2 other papers
2 later
1812.04948 (2018-12)
A Style-Based Generator Architecture for Generative Adversarial Networks
2605.11983 (2026-05)
QDSB: Quantized Diffusion Schrödinger Bridges
avg_losses 1 other paper
1 dated earlier
1511.05644 (2015-11)
Adversarial Autoencoders
make_image_grid 1 other paper
1 dated earlier
1511.05644 (2015-11)
Adversarial Autoencoders

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

Autoencoders provide a powerful framework for learning compressed representations by encoding all of the information needed to reconstruct a data point in a latent code. In some cases, autoencoders can "interpolate": By decoding the convex combination of the latent codes for two datapoints, the autoencoder can produce an output which semantically mixes characteristics from the datapoints. In this paper, we propose a regularization procedure which encourages interpolated outputs to appear more realistic by fooling a critic network which has been trained to recover the mixing coefficient from interpolated data. We then develop a simple benchmark task where we can quantitatively measure the extent to which various autoencoders can interpolate and show that our regularizer dramatically improves interpolation in this setting. We also demonstrate empirically that our regularizer produces latent codes which are more effective on downstream tasks, suggesting a possible link between interpolation abilities and learning useful representations.

For agents

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

get_harvested_code_for_paper("1807.07543")
get_code_for_paper("1807.07543")
have("1807.07543")

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

curl https://syntology.ai/api/ran/1807.07543.json

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

[![Syntology run record](https://syntology.ai/api/ran/1807.07543.svg)](https://syntology.ai/paper/1807.07543)

Connect an agent — have() is free.