SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2311.17643 · 2023

Thera: Aliasing-Free Arbitrary-Scale Super-Resolution with Neural Heat Fields

arXiv · PDF · Open in the Atlas

Code that ran

9 of 10 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (1 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 8 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/2311.17643.json · badge: /api/ran/2311.17643.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
prs-eth/thera canonical (classifier label)
classifier basis: census
9 ran · 1 no recorded run
FunctionStatusWhere it lives
Dense Ran
ran; our contract draft was wrong, not the code · constructed an object; no result was computed
prs-eth/thera/model/swin_ir.py
code served (permissive licence) · get_code("eaa96cc883399233")
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 an instance of its own class, Dense: the run constructed the object and did not call it, so no result was computed. 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 jax
import jax.numpy as jnp
import flax.linen as nn
import math
from jaxtyping import Array

def trunc_normal(mean=0.0, std=1.0, a=-2.0, b=2.0, dtype=jnp.float32):
    """Truncated normal initialization function"""
    def init(key, shape, dtype=dtype) -> Array:
        def norm_cdf(x):
            return (1.0 + math.erf(x / math.sqrt(2.0))) / 2.0
        l = norm_cdf((a - mean) / std)
        u = norm_cdf((b - mean) / std)
        out = jax.random.uniform(key, shape, dtype=dtype, minval=2 * l - 1, maxval=2 * u - 1)
        out = jax.scipy.special.erfinv(out) * std * math.sqrt(2.0) + mean
        return jnp.clip(out, a, b)
    return init

def Dense(features, use_bias=True, kernel_init=trunc_normal(std=0.02), bias_init=nn.initializers.zeros):
    return nn.Dense(features, use_bias=use_bias, kernel_init=kernel_init, bias_init=bias_init)

ARGS = {
    "features": 8,
    "use_bias": True,
    "kernel_init": trunc_normal(std=0.02),
    "bias_init": nn.initializers.zeros
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
flax, jax, jaxtyping — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Dense — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Dense 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
build_tail Ran
ran; our contract draft was wrong, not the code
prs-eth/thera/model/tail.py
code served (permissive licence) · get_code("8d339b80b80b87ed")
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 function. 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)

class ConvNeXt(nn.Module):
    def __init__(self, blocks):
        super().__init__()
        self.blocks = blocks
    
    def forward(self, x):
        return x

class SwinIR(nn.Module):
    def __init__(self, depths, num_heads):
        super().__init__()
        self.depths = depths
        self.num_heads = num_heads
    
    def forward(self, x):
        return x

ARGS = {
    'size': 'air'
}
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 function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
convert_edsr_checkpoint Ran
ran; our contract draft was wrong, not the code
prs-eth/thera/model/edsr.py
code served (permissive licence) · get_code("9a26b082285c928a")
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 FrozenDict. 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 numpy as np
import jax.numpy as jnp
from flax.core.frozen_dict import freeze

torch.manual_seed(0)

# Build a torch_dict that mimics an EDSR checkpoint structure
# The function expects keys like:
# - 'weight', 'bias' at top level (layer leaves)
# - Numeric keys like '0', '1', etc. for layer indices
# - Nested structures like '0.weight', '0.bias', etc.

torch_dict = {
    # Top-level leaves
    'weight': torch.randn(3, 4, 3, 3),  # Conv weight: (out_channels, in_channels, H, W)
    'bias': torch.randn(3),             # Conv bias
    
    # Numeric layers (will be converted to 'layers_0', 'layers_1', etc.)
    '0.weight': torch.randn(4, 3, 3, 3),
    '0.bias': torch.randn(4),
    
    '1.weight': torch.randn(5, 4, 3, 3),
    '1.bias': torch.randn(5),
    
    # Non-numeric top-level keys
    'head.weight': torch.randn(4, 3, 3, 3),
    'head.bias': torch.randn(4),
    
    'tail.weight': torch.randn(3, 4, 3, 3),
    'tail.bias': torch.randn(3),
    
    'add_mean': torch.tensor([0.485, 0.456, 0.406]),
    'sub_mean': torch.tensor([0.229, 0.224, 0.225]),
}

ARGS = {
    'torch_dict': torch_dict,
    'no_upsampling': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
flax, jax, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type FrozenDict — a FrozenDict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a FrozenDict 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
linear_up Ran
ran; our contract draft was wrong, not the code
prs-eth/thera/model/init.py
code served (permissive licence) · get_code("6eea895669cb9dd3")
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 function. 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 jax
import jax.numpy as jnp
from jaxtyping import Array

# The function linear_up takes a single parameter: scale (float)
# It returns a function (init) that can be used as an initializer
# The init function takes key, shape, and dtype parameters

# scale: a float value that controls the scaling of the initialization
ARGS = {
    "scale": 0.5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, jaxtyping — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
numpy_collate Ran
ran; our contract draft was wrong, not the code
prs-eth/thera/data.py
code served (permissive licence) · get_code("505b6e225f0f9008")
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 list. 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 numpy as np
import torch
from torch.utils.data import Dataset, default_collate

torch.manual_seed(0)

class SimpleDataset(Dataset):
    def __init__(self, size=2):
        self.size = size
    
    def __len__(self):
        return self.size
    
    def __getitem__(self, idx):
        x = torch.randn(4, 3)
        y = torch.tensor(idx, dtype=torch.long)
        return x, y

dataset = SimpleDataset(size=2)
batch = [dataset[i] for i in range(len(dataset))]

ARGS = {
    "batch": batch
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list — a list 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
open_img Ran
ran, honoured the contract we drafted
prs-eth/thera/data.py
code served (permissive licence) · get_code("ab6ae61c2be5a832")
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 3, dtype family int, shape (3, 4, 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
import torch
import torch.nn.functional as f
from PIL import Image
import io

torch.manual_seed(0)

def to_uint8_tensor(image: Image):
    return torch.from_numpy(np.array(image)).permute(2, 0, 1)

def open_img(path):
    with open(str(path), 'rb') as f:
        img = Image.open(f).convert('RGB')
        img = to_uint8_tensor(img)
    return img

# Create a small test image in memory
img_array = np.random.randint(0, 256, (4, 3, 3), dtype=np.uint8)
pil_img = Image.fromarray(img_array, mode='RGB')

# Save to a temporary file path for testing
import tempfile
import os

temp_dir = tempfile.mkdtemp()
temp_path = os.path.join(temp_dir, "test_image.png")
pil_img.save(temp_path)

ARGS = {
    "path": temp_path
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, 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 3, dtype family int, shape (3, 4, 3) — shape (3, 4, 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
to_uint8_tensor Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
prs-eth/thera/data.py
code served (permissive licence) · get_code("3d6492c948231fd9")
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 3, dtype family int, shape (3, 3, 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
import torch
from PIL import Image

torch.manual_seed(0)

# Create a small PIL Image with shape (height=3, width=4, channels=3)
# to have distinct dimensions
image_array = np.random.randint(0, 256, (3, 4, 3), dtype=np.uint8)
pil_image = Image.fromarray(image_array, mode='RGB')

ARGS = {
    'image': pil_image
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
PIL, 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 3, dtype family int, shape (3, 3, 4) — shape (3, 3, 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
trunc_normal Ran
ran; our contract draft was wrong, not the code
prs-eth/thera/model/swin_ir.py
code served (permissive licence) · get_code("e07071f77c4d8d4d")
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 function. 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 jax
import jax.numpy as jnp
import math

# Set random seed for reproducibility
jax.config.update("jax_default_prng_impl", "threefry2x32")

ARGS = {
    "mean": 0.0,
    "std": 1.0,
    "a": -2.0,
    "b": 2.0,
    "dtype": jnp.float32,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, jaxtyping — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
uniform_between Ran
ran; our contract draft was wrong, not the code
prs-eth/thera/model/init.py
code served (permissive licence) · get_code("d19a418cbb8a5b03")
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 function. 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 jax
import jax.numpy as jnp

ARGS = {
    "a": 0.0,
    "b": 1.0,
    "dtype": jnp.float32,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
jax, jaxtyping — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type function — a function output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a function 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
window_partition Not yet run
unverified (harvested, no recorded run)
prs-eth/thera/model/swin_ir.py
code served (permissive licence) · get_code("7f0d8d7253a371eb")
How it ran

no invocation recorded.

Also present in other papers' code

Of 10 samples lifted from this paper's linked repositories, 10 carry a function identity and 0 do not. Of the 10 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 10 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

Recent approaches to arbitrary-scale single image super-resolution (ASR) use neural fields to represent continuous signals that can be sampled at arbitrary resolutions. However, point-wise queries of neural fields do not naturally match the point spread function (PSF) of pixels, which may cause aliasing in the super-resolved image. Existing methods attempt to mitigate this by approximating an integral version of the field at each scaling factor, compromising both fidelity and generalization. In this work, we introduce neural heat fields, a novel neural field formulation that inherently models a physically exact PSF. Our formulation enables analytically correct anti-aliasing at any desired output resolution, and -- unlike supersampling -- at no additional cost. Building on this foundation, we propose Thera, an end-to-end ASR method that substantially outperforms existing approaches, while being more parameter-efficient and offering strong theoretical guarantees. The project page is at https://therasr.github.io.

For agents

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

get_harvested_code_for_paper("2311.17643")
get_code_for_paper("2311.17643")
have("2311.17643")

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

curl https://syntology.ai/api/ran/2311.17643.json

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

[![Syntology run record](https://syntology.ai/api/ran/2311.17643.svg)](https://syntology.ai/paper/2311.17643)

Connect an agent — have() is free.