SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1808.08718 · 2018

Wide Activation for Efficient and Accurate Image Super-Resolution

arXiv · PDF · Open in the Atlas

Code that ran

9 of 17 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 9 have no contract verdict; 0 failures of our instrument; 8 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/1808.08718.json · badge: /api/ran/1808.08718.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
krasserm/super-resolution community (classifier label)
classifier basis not recorded on the link
6 ran · 2 no recorded run
epochlab/xres community (classifier label)
classifier basis not recorded on the link
3 ran · 4 no recorded run
SJHNJU/WDSR reimplementation (classifier label)
classifier basis: classifier_2025
0 ran · 2 no recorded run
FunctionStatusWhere it lives
build_srgan Ran
ran on a synthesized input
epochlab/xres/model/srgan.py
code served (permissive licence) · get_code("7e396aede5e06e28")
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 Functional. 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
from tensorflow.keras import Input, Model
from tensorflow.keras.layers import Conv2D, BatchNormalization, UpSampling2D, Activation, LeakyReLU, PReLU, Add, Dense, Flatten

# Set seed for reproducibility
tf.random.set_seed(0)

# The function build_srgan takes one parameter: input_shape
# input_shape should be a tuple representing (height, width, channels)
# Based on the function body, it's used to create an Input layer
# The Conv2D operations and other layers suggest this should be a 3D shape (height, width, channels)

ARGS = {
    'input_shape': (4, 3, 3)  # (height=4, width=3, channels=3) - small tensor dimensions, distinct sizes
}
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 Functional — a Functional output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a Functional 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
load_image Ran
ran on a synthesized input
krasserm/super-resolution/utils.py
code served (permissive licence) · get_code("0ab4dee93051e84f")
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 int, 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 numpy as np
from PIL import Image
import tempfile
import os

# Create a temporary image file
temp_dir = tempfile.mkdtemp()
temp_image_path = os.path.join(temp_dir, "test_image.png")

# Create a small test image and save it
test_image_array = np.random.randint(0, 256, (4, 3, 3), dtype=np.uint8)
test_image = Image.fromarray(test_image_array)
test_image.save(temp_image_path)

ARGS = {
    "path": temp_image_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, 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 int, 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
log_callback Ran
ran on a synthesized input
epochlab/xres/utils.py
code served (permissive licence) · get_code("ddec0ea1ef169791")
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 os
import tempfile
import tensorflow as tf
from tensorflow.keras import layers

# Create minimal generator and discriminator models
class Generator(tf.keras.Model):
    def __init__(self):
        super().__init__()
        self.dense = layers.Dense(10)
    
    def call(self, inputs):
        return self.dense(inputs)

class Discriminator(tf.keras.Model):
    def __init__(self):
        super().__init__()
        self.dense = layers.Dense(1)
    
    def call(self, inputs):
        return self.dense(inputs)

# Create a temporary directory for output
outdir = tempfile.mkdtemp()

# Create model instances
generator = Generator()
discriminator = Discriminator()

# Create optimizers
generator_optimizer = tf.keras.optimizers.Adam(learning_rate=0.0002)
discriminator_optimizer = tf.keras.optimizers.Adam(learning_rate=0.0002)

# Define ARGS dictionary with all parameters
ARGS = {
    'outdir': outdir,
    'generator': generator,
    'discriminator': discriminator,
    'generator_optimizer': generator_optimizer,
    'discriminator_optimizer': discriminator_optimizer,
}
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, 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
random_crop Ran
ran on a synthesized input
krasserm/super-resolution/data.py
code served (permissive licence) · get_code("40c578a7c8a918e6")
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
import numpy as np

# Create small test images
# LR image: 100x100 (low resolution)
# HR image: 200x200 (high resolution, 2x upscaled)
lr_img = tf.constant(np.random.randint(0, 256, (100, 100, 3), dtype=np.uint8), dtype=tf.uint8)
hr_img = tf.constant(np.random.randint(0, 256, (200, 200, 3), dtype=np.uint8), dtype=tf.uint8)

ARGS = {
    "lr_img": lr_img,
    "hr_img": hr_img,
    "hr_crop_size": 96,
    "scale": 2,
}
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 — 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
random_flip Ran
ran on a synthesized input
krasserm/super-resolution/data.py
code served (permissive licence) · get_code("256bed53fab6520f")
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

# Create small test images for lr_img and hr_img
# lr_img: low-resolution image, shape (batch, height, width, channels)
# hr_img: high-resolution image, shape (batch, height, width, channels)
lr_img = tf.constant([[[[1.0, 2.0, 3.0],
                        [4.0, 5.0, 6.0],
                        [7.0, 8.0, 9.0],
                        [10.0, 11.0, 12.0]],
                       [[13.0, 14.0, 15.0],
                        [16.0, 17.0, 18.0],
                        [19.0, 20.0, 21.0],
                        [22.0, 23.0, 24.0]]]], dtype=tf.float32)

hr_img = tf.constant([[[[1.0, 2.0],
                        [3.0, 4.0]],
                       [[5.0, 6.0],
                        [7.0, 8.0]]]], dtype=tf.float32)

ARGS = {
    'lr_img': lr_img,
    'hr_img': hr_img
}
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
random_rotate Ran
ran on a synthesized input
krasserm/super-resolution/data.py
code served (permissive licence) · get_code("4cb58354ae7610d2")
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
import numpy as np

# Create small sample images for low-resolution and high-resolution
# lr_img: low-resolution image, shape (height, width, channels)
# hr_img: high-resolution image, shape (height, width, channels)
lr_img = tf.constant(np.random.rand(4, 3, 3).astype(np.float32))
hr_img = tf.constant(np.random.rand(8, 6, 3).astype(np.float32))

ARGS = {
    "lr_img": lr_img,
    "hr_img": hr_img,
}
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 — 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
resolve Ran
ran on a synthesized input
krasserm/super-resolution/model/common.py
code served (permissive licence) · get_code("fcb21bb87592664b")
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 int, shape (2, 4, 3, 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 tensorflow as tf
import numpy as np

# Create a minimal model that takes a low-resolution batch and outputs super-resolved images
class SimpleSuperResolutionModel(tf.Module):
    def __init__(self):
        super().__init__()
        self.dense = tf.keras.layers.Dense(12, activation='relu')
        self.out = tf.keras.layers.Dense(12)
    
    def __call__(self, x):
        # x shape: (batch, height, width, channels)
        # Flatten, process, and reshape back
        batch_size = tf.shape(x)[0]
        original_shape = tf.shape(x)
        
        x_flat = tf.reshape(x, [batch_size, -1])
        x = self.dense(x_flat)
        x = self.out(x)
        
        # Reshape to match expected output shape (same as input for simplicity)
        sr = tf.reshape(x, original_shape)
        return sr

# Create the model
model = SimpleSuperResolutionModel()

# Create a small low-resolution batch
# Shape: (batch_size=2, height=4, width=3, channels=1)
# Using float32 as the function casts to float32 anyway
lr_batch = tf.constant(
    np.random.randint(0, 100, size=(2, 4, 3, 1), dtype=np.int32),
    dtype=tf.int32
)

ARGS = {
    "model": model,
    "lr_batch": lr_batch
}
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 int, shape (2, 4, 3, 1) — shape (2, 4, 3, 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
resolve_single Ran
ran on a synthesized input
krasserm/super-resolution/model/common.py
code served (permissive licence) · get_code("ae235fb465d0b309")
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 3, dtype family int, shape (2, 2, 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 tensorflow as tf
import numpy as np

# Create a minimal model that takes low-res input and outputs super-resolved output
class SimpleUpscaleModel(tf.Module):
    def __init__(self):
        super(SimpleUpscaleModel, self).__init__()
        self.dense = tf.keras.layers.Dense(12, activation='relu')
        self.output_layer = tf.keras.layers.Dense(12)
    
    def __call__(self, lr_batch):
        # lr_batch shape: (batch, height, width, channels) -> (batch, 12)
        x = tf.reshape(lr_batch, [tf.shape(lr_batch)[0], -1])
        x = self.dense(x)
        # Output should be in range [0, 255] after clipping and rounding
        x = self.output_layer(x)
        # Reshape back to image-like format
        x = tf.reshape(x, [tf.shape(lr_batch)[0], 2, 2, 3])
        return x

# Instantiate the model
model = SimpleUpscaleModel()

# Create a low-res input image (batch_size=1, height=2, width=2, channels=3)
# Values in uint8 range for realism
lr = tf.constant(np.random.randint(0, 256, (2, 2, 3), dtype=np.uint8), dtype=tf.uint8)

ARGS = {
    "model": model,
    "lr": lr
}
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 3, dtype family int, shape (2, 2, 3) — shape (2, 2, 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
upsampling_block Ran
ran on a synthesized input
epochlab/xres/model/edsr.py
code served (permissive licence) · get_code("72ae868bbca101e0")
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 KerasTensor. 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

# Create a minimal input tensor for the model
# The function expects a model/tensor input that will be passed through Conv2D, UpSampling2D, PReLU
# Shape: (batch_size, height, width, channels)
input_tensor = tf.keras.Input(shape=(4, 4, 3))

ARGS = {
    'model': input_tensor,
    'num_filters': 32,
    'kernal_size': 3,
    'strides': 1
}
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 KerasTensor — a KerasTensor 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
build_vgg Not yet run
unverified (harvested, no recorded run)
epochlab/xres/model/common.py
code served (permissive licence) · get_code("043379256a5acbc8")
How it ran

no invocation recorded.

edsr_residual Not yet run
unverified (harvested, no recorded run)
epochlab/xres/model/edsr.py
code served (permissive licence) · get_code("d798286016d72cbd")
How it ran

no invocation recorded.

evaluate Not yet run
unverified (harvested, no recorded run)
krasserm/super-resolution/model/common.py
code served (permissive licence) · get_code("e0f9b50298061760")
How it ran

no invocation recorded.

gridless_upsampling Not yet run
unverified (harvested, no recorded run)
epochlab/xres/model/edsr.py
code served (permissive licence) · get_code("8db36c7033b552cb")
How it ran

no invocation recorded.

psnr Not yet run
unverified (harvested, no recorded run)
SJHNJU/WDSR/psnr.py
pointer only (licence: NONE) · get_code("5065f8eb4503235e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, SJHNJU/WDSR 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. No invocation recorded.

res_block Not yet run
unverified (harvested, no recorded run)
krasserm/super-resolution/model/edsr.py
code served (permissive licence) · get_code("71ba60506facd18f")
How it ran

no invocation recorded.

residual_block Not yet run
unverified (harvested, no recorded run)
epochlab/xres/model/srgan.py
code served (permissive licence) · get_code("61df72fd2cd3dba5")
How it ran

no invocation recorded.

test Not yet run
unverified (harvested, no recorded run)
SJHNJU/WDSR/psnr.py
pointer only (licence: NONE) · get_code("37c6b07e255ca382")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, SJHNJU/WDSR 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. No invocation recorded.

Also present in other papers' code

Of 17 samples lifted from this paper's linked repositories, 17 carry a function identity and 0 do not. Of the 17 distinct functions with an identity, 8 are also present in other papers whose code we hold (8 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.

FunctionAlso present inEarliest-datedOthers named
evaluate 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
load_image 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
random_crop 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
random_flip 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
random_rotate 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
res_block 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
resolve 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks
resolve_single 1 other paper
1 dated earlier
1602.07868 (2016-02)
Weight Normalization: A Simple Reparameterization to Accelerate Training of Deep Neural Networks

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

In this report we demonstrate that with same parameters and computational budgets, models with wider features before ReLU activation have significantly better performance for single image super-resolution (SISR). The resulted SR residual network has a slim identity mapping pathway with wider (\(2\times\) to \(4\times\)) channels before activation in each residual block. To further widen activation (\(6\times\) to \(9\times\)) without computational overhead, we introduce linear low-rank convolution into SR networks and achieve even better accuracy-efficiency tradeoffs. In addition, compared with batch normalization or no normalization, we find training with weight normalization leads to better accuracy for deep super-resolution networks. Our proposed SR network \textit{WDSR} achieves better results on large-scale DIV2K image super-resolution benchmark in terms of PSNR with same or lower computational complexity. Based on WDSR, our method also won 1st places in NTIRE 2018 Challenge on Single Image Super-Resolution in all three realistic tracks. Experiments and ablation studies support the importance of wide activation for image super-resolution. Code is released at: https://github.com/JiahuiYu/wdsr_ntire2018

For agents

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

get_harvested_code_for_paper("1808.08718")
get_code_for_paper("1808.08718")
have("1808.08718")

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

curl https://syntology.ai/api/ran/1808.08718.json

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

[![Syntology run record](https://syntology.ai/api/ran/1808.08718.svg)](https://syntology.ai/paper/1808.08718)

Connect an agent — have() is free.