SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2204.06125 · arXiv.org · 2022

Hierarchical Text-Conditional Image Generation with CLIP Latents

Casey Chu, Aditya Ramesh, Mark Chen, Alex Nichol, Prafulla Dhariwal

arXiv · PDF · Open in the Atlas

Code that ran

33 of 38 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (7 of them constructed an object rather than computing a result; 3 honoured the contract we drafted, 6 violated it and 13 have no contract verdict; 11 failures of our instrument; 5 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/2204.06125.json · badge: /api/ran/2204.06125.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
liyinqi/un2clip community (classifier label)
classifier basis not recorded on the link
15 ran · 2 no recorded run
harvard-visionlab/lrm-steering community
Papers with Code archive
10 ran · 3 no recorded run
lucidrains/DALLE2-pytorch community
Papers with Code archive
5 ran · 0 no recorded run
laion-ai/conditioned-prior community (classifier label)
classifier basis not recorded on the link
2 ran · 0 no recorded run
copy not recorded community
not in the paper's declared repositories
1 ran · 0 no recorded run
FunctionStatusWhere it lives
AdaptiveFullstackNorm Ran
ran on a synthesized input
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("126ed34c3a58c622")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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 torch
import torch.nn as nn
import torch.nn.functional as F
from torch import Tensor

class AdaptiveFullstackNorm(nn.LayerNorm):

    def forward(self, x: Tensor, *args) -> Tensor:
        memory_format = torch.channels_last if x.is_contiguous(memory_format=torch.channels_last) else torch.contiguous_format
        weight = F.interpolate(self.weight.unsqueeze(0), size=x.shape[-2:], mode='bilinear', align_corners=True).squeeze()
        bias = F.interpolate(self.bias.unsqueeze(0), size=x.shape[-2:], mode='bilinear', align_corners=True).squeeze()
        x = F.layer_norm(x, x.shape[-3:], weight, bias, self.eps)
        x = x.to(memory_format=memory_format)
        return x


INIT_ARGS = {
    'normalized_shape': (4, 3, 2),
    'eps': 1e-5,
    'elementwise_affine': True
}

ARGS = {
    'x': torch.randn(2, 4, 3, 2)
}

if __name__ == '__main__':
    module = AdaptiveFullstackNorm(**INIT_ARGS)
    output = module(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    print(f"Output shape: {output.shape}")
    print(f"Output:\n{output}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
AdaptiveUpsample Ran
ran on a synthesized input · constructed an object; no result was computed
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("8c00bf2ba3f33794")
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, AdaptiveUpsample: 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Create a small input tensor for testing
# Shape: (batch=2, channels=4, height=8, width=8)
x = torch.randn(2, 4, 8, 8)

# Define output size for upsampling
out_size = (16, 16)

# Map all parameters of AdaptiveUpsample.__init__
ARGS = {
    'upsample_mode': 'AdaptiveAvgPool2d'
}
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 AdaptiveUpsample — 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 AdaptiveUpsample 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
AddSpatialDimension Ran
ran on a synthesized input
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("8aae4fa13c828688")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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 torch
import torch.nn as nn
from torch import Tensor

class MemoryFormatModule(nn.Module):

    def __init__(self):
        super().__init__()
        self.memory_format = torch.contiguous_format

    def set_memory_format(self, memory_format):
        self.memory_format = memory_format

    def _apply(self, fn):
        """
            When we call model.to(memory_format=...) we want to capture the memory_format 
            and apply it to the tensor `x` in the forward pass after it's unsqueezed to add 
            the spatial dimensions.
            
            This solution uses introspection on Python functions (specifically, looking into 
            the closure of the convert function). While this works for the current version of 
            PyTorch I'm using (1.13.1.post200), it does depend on internal details that might 
            change in future versions.             
        """
        super(MemoryFormatModule, self)._apply(fn)
        if fn.__name__ == 'convert':
            closure_vars = {k: v for k, v in zip(fn.__code__.co_freevars, (c.cell_contents for c in fn.__closure__))}
            memory_format = closure_vars.get('convert_to_format', None)
            if memory_format is not None:
                self.set_memory_format(memory_format)
        return self

class AddSpatialDimension(MemoryFormatModule):

    def forward(self, x: Tensor, *args) -> Tensor:
        if len(x.shape) == 2:
            x = x.unsqueeze(-1).unsqueeze(-1)
            x = x.to(memory_format=self.memory_format)
        return x

    def __repr__(self):
        return f'{self.__class__.__name__}()'


INIT_ARGS = {}

ARGS = {
    'x': torch.randn(2, 4)
}

if __name__ == '__main__':
    model = AddSpatialDimension(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    print(f"Output shape: {output.shape}")
    print(f"Output: {output}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
ChannelNorm Ran
ran on a synthesized input
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("753410ece03ec32e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, 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 torch
import torch.nn as nn
import torch.nn.functional as F
from torch import Tensor

class ChannelNorm(nn.LayerNorm):

    def forward(self, x: Tensor, *args) -> Tensor:
        if len(x.shape) == 4:
            x = x.permute(0, 2, 3, 1)
            x = F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps)
            x = x.permute(0, 3, 1, 2)
        elif len(x.shape) == 2:
            x = F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps)
        else:
            raise ValueError(f'expected x.shape to be 2 or 4 numbers, got {x.shape}')
        return x


INIT_ARGS = {
    'normalized_shape': (4,),
    'eps': 1e-5,
    'elementwise_affine': True
}

ARGS = {
    'x': torch.randn(2, 4, 3, 5)
}

if __name__ == '__main__':
    model = ChannelNorm(**INIT_ARGS)
    output = model(**ARGS)
    print(f"Input shape: {ARGS['x'].shape}")
    print(f"Output shape: {output.shape}")
    print(f"Output:\n{output}")
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
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 — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
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: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
FeatureExtractor Ran
ran on a synthesized input · constructed an object; no result was computed
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("a7f6adc337b5c0ac")
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, FeatureExtractor: 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 torch
import torch.nn as nn

torch.manual_seed(0)

# Create a small model with named modules for testing
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.layer1 = nn.Linear(4, 3)
        self.layer2 = nn.Linear(3, 2)
    
    def forward(self, x):
        x = self.layer1(x)
        x = self.layer2(x)
        return x

model = SimpleModel()

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

# Define ARGS dictionary with all required parameters
ARGS = {
    'model': model,
    'layers': ['layer1', 'layer2'],
    'retain': True,
    'detach': True,
    'clone': True,
    'device': 'cpu'
}
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 FeatureExtractor — the run constructed an instance of its own class and did not call it; no shape 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
FeedbackScale Ran
ran on a synthesized input · constructed an object; no result was computed
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("3d2b010f98ba752d")
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, FeedbackScale: 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'mode': 'tanh'
}
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 FeedbackScale — 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 FeedbackScale 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
MemoryFormatModule Ran
ran on a synthesized input · constructed an object; no result was computed
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("11fbbabed042993e")
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, MemoryFormatModule: 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 torch
import torch.nn as nn

torch.manual_seed(0)

class MemoryFormatModule(nn.Module):

    def __init__(self):
        super().__init__()
        self.memory_format = torch.contiguous_format

    def set_memory_format(self, memory_format):
        self.memory_format = memory_format

    def _apply(self, fn):
        """
            When we call model.to(memory_format=...) we want to capture the memory_format 
            and apply it to the tensor `x` in the forward pass after it's unsqueezed to add 
            the spatial dimensions.
            
            This solution uses introspection on Python functions (specifically, looking into 
            the closure of the convert function). While this works for the current version of 
            PyTorch I'm using (1.13.1.post200), it does depend on internal details that might 
            change in future versions.             
        """
        super(MemoryFormatModule, self)._apply(fn)
        if fn.__name__ == 'convert':
            closure_vars = {k: v for k, v in zip(fn.__code__.co_freevars, (c.cell_contents for c in fn.__closure__))}
            memory_format = closure_vars.get('convert_to_format', None)
            if memory_format is not None:
                self.set_memory_format(memory_format)
        return self

# The __init__ method takes no parameters, so we just need to instantiate it
model = MemoryFormatModule()

ARGS = {}
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 MemoryFormatModule — 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 MemoryFormatModule 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
NoiseScheduler Ran
ran on a synthesized input · constructed an object; no result was computed
lucidrains/DALLE2-pytorch/dalle2_pytorch/dalle2_pytorch.py
code served (permissive licence) · get_code("78cd7549fb76becd")
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, NoiseScheduler: 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 torch
import torch.nn as nn

torch.manual_seed(0)

ARGS = {
    'beta_schedule': 'cosine',
    'timesteps': 100,
    'loss_type': 'l2',
    'p2_loss_weight_gamma': 0.0,
    'p2_loss_weight_k': 1
}
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 NoiseScheduler — 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 NoiseScheduler 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
NormSquashResize Ran
ran on a synthesized input · constructed an object; no result was computed
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("5a6abd14dc0a7611")
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, NormSquashResize: 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 torch
import torch.nn as nn

torch.manual_seed(0)

# The function signature is:
# NormSquashResize(in_channels, in_shape, out_shape, norm_type='ChannelNorm', scale_type='tanh', resize_type='UpsampleBilinear')

ARGS = {
    'in_channels': 4,
    'in_shape': 8,
    'out_shape': 16,
    'norm_type': 'ChannelNorm',
    'scale_type': 'tanh',
    'resize_type': 'UpsampleBilinear'
}
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 NormSquashResize — 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 NormSquashResize 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
Normalize Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/modules/diffusionmodules/model.py
code served (permissive licence) · get_code("9fcdaa6e423e8aa7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned GroupNorm. 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

torch.manual_seed(0)

# The Normalize function takes in_channels and num_groups (with default 32)
# It returns a GroupNorm layer

# in_channels: number of channels for the GroupNorm layer
# num_groups: number of groups to separate the channels into (default 32)

# We need to provide valid arguments where num_groups divides in_channels evenly
# Let's use in_channels=64 so it's divisible by the default num_groups=32

ARGS = {
    'in_channels': 64,
    'num_groups': 32
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type GroupNorm — a GroupNorm output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a GroupNorm 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
ResizeNormSquash Ran
ran on a synthesized input · constructed an object; no result was computed
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("67f277bd7c489e7e")
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, ResizeNormSquash: 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 torch
import torch.nn as nn
from torch.nn.modules.utils import _pair

torch.manual_seed(0)

ARGS = {
    'in_channels': 4,
    'in_shape': 8,
    'out_shape': 16,
    'norm_type': 'ChannelNorm',
    'scale_type': 'tanh',
    'resize_type': 'UpsampleBilinear'
}
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 ResizeNormSquash — 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 ResizeNormSquash 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
append_dims Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
liyinqi/un2clip/ldm/models/diffusion/sampling_util.py
code served (permissive licence) · get_code("082e51fd548d78b2")
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 5, dtype family float, shape (2, 4, 3, 1, 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 torch

torch.manual_seed(0)

# The function append_dims takes a tensor x and appends dimensions to it
# until it has target_dims total dimensions.
# It uses:
# - x.ndim (attribute access)
# - x[(...,) + (None,) * dims_to_append] (indexing operation)

# Create a small tensor with a few dimensions
x = torch.randn(2, 4, 3)  # 3D tensor with shape (2, 4, 3)

# target_dims should be an integer greater than x.ndim
target_dims = 5

ARGS = {
    "x": x,
    "target_dims": target_dims
}
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 5, dtype family float, shape (2, 4, 3, 1, 1) — shape (2, 4, 3, 1, 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
autocast Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/util.py
code served (permissive licence) · get_code("32af1c12fd826ef2")
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. 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.nn as nn

torch.manual_seed(0)

# The autocast function takes a single parameter: a function f
# It returns a wrapped version of that function with autocast enabled

# Create a simple function to wrap
def sample_forward(x):
    """A simple function that does some computation"""
    return x * 2 + 1

ARGS = {
    "f": sample_forward
}
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
clip_fwd Ran
ran on a synthesized input
liyinqi/un2clip/eval_mmvpvlm.py
code served (permissive licence) · get_code("2b22d96b91008d68")
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 3, 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.nn as nn

torch.manual_seed(0)

class MockCLIPModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, imgs, text):
        # Return (logits_per_image, logits_per_text)
        batch_size = imgs.shape[0]
        logits_per_image = self.linear(imgs)  # (batch, 3)
        logits_per_text = torch.randn(batch_size, batch_size)  # (batch, batch)
        return logits_per_image, logits_per_text

model = MockCLIPModel()
imgs = torch.randn(2, 4)
text1 = torch.randn(2, 4)
text2 = torch.randn(2, 4)

ARGS = {
    "model": model,
    "imgs": imgs,
    "text1": text1,
    "text2": text2
}
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, ndim 3, 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
cosine_beta_schedule Ran
ran, honoured the contract we drafted
lucidrains/DALLE2-pytorch/dalle2_pytorch/dalle2_pytorch.py
code served (permissive licence) · get_code("63d04624697c8f60")
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 (10,) 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)

ARGS = {
    "timesteps": 10,
    "s": 0.008,
}
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 1, dtype family float, shape (10,) — shape (10,) 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
default Ran
ran, violated the contract we drafted
liyinqi/un2clip/ldm/modules/attention.py
code served (permissive licence) · get_code("424012cb37b31172")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned int, ndim 0, dtype family int, 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.

from inspect import isfunction

def exists(val):
    return val is not None

def default(val, d):
    if exists(val):
        return val
    return d() if isfunction(d) else d

# Test case 1: val exists (not None), d is a regular value
ARGS_1 = {
    'val': 42,
    'd': 100
}

# Test case 2: val is None, d is a regular value
ARGS_2 = {
    'val': None,
    'd': 100
}

# Test case 3: val is None, d is a function
def default_func():
    return 200

ARGS_3 = {
    'val': None,
    'd': default_func
}

# Test case 4: val exists, d is a function
ARGS_4 = {
    'val': "existing_value",
    'd': lambda: "default_value"
}

# Use one of the test cases as the final ARGS
ARGS = ARGS_3
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type int, ndim 0, dtype family int, 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
default Ran
ran, violated the contract we drafted
this paper's copy was not recorded; identical code first harvested from lucidrains/flash-cosine-sim-attention
pointer only · get_code("915fbeb88e13d4a3")
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.

disabled_train Ran
ran, violated the contract we drafted
liyinqi/un2clip/ldm/models/diffusion/ddpm.py
code served (permissive licence) · get_code("4cb732f513d69dfd")
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) and returned SimpleModel. 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 SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
    
    def forward(self, x):
        return self.linear(x)

model = SimpleModel()

ARGS = {
    "self": model,
    "mode": True
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type SimpleModel — a SimpleModel output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModel 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
exists Ran
ran, violated the contract we drafted
liyinqi/un2clip/ldm/modules/attention.py
pointer only (licence: MIT); the stored copy get_code serves, from ThomasMrY/VCT, is pointer only (licence: NONE) · get_code("aa5486a3650902d8")
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 ThomasMrY/VCT, 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.

first Ran
ran, violated the contract we drafted
lucidrains/DALLE2-pytorch/dalle2_pytorch/dalle2_pytorch.py
code served (permissive licence) · get_code("3072294c9b32cf4b")
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 int, ndim 0, dtype family int, 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)

# The function `first` takes an array-like object and an optional default value
# It returns the first element of the array, or the default if the array is empty

# Test with a simple list
arr = [1, 2, 3, 4, 5]
d = None

ARGS = {
    "arr": arr,
    "d": d
}
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 int, ndim 0, dtype family int, 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_layer_shapes Ran
ran; our contract draft was wrong, not the code
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("c32223f1f7f06e58")
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 dict. 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)

# Create a simple model with named modules for testing
class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.conv1 = nn.Conv2d(3, 8, kernel_size=3, padding=1)
        self.relu1 = nn.ReLU()
        self.conv2 = nn.Conv2d(8, 16, kernel_size=3, padding=1)
        self.relu2 = nn.ReLU()
        self.fc = nn.Linear(16 * 4 * 4, 10)
    
    def forward(self, x):
        x = self.conv1(x)
        x = self.relu1(x)
        x = self.conv2(x)
        x = self.relu2(x)
        x = x.view(x.size(0), -1)
        x = self.fc(x)
        return x

model = SimpleModel()
model.eval()

# Create a small batch of input data
# batch_size=2, channels=3, height=4, width=4
x = torch.randn(2, 3, 4, 4)

# Define layer names to extract features from
layer_names = ['relu1', 'relu2']

ARGS = {
    'model': model,
    'layer_names': layer_names,
    'x': x
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
get_timestep_embedding Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
liyinqi/un2clip/ldm/modules/diffusionmodules/model.py
code served (permissive licence) · get_code("cb49209c125de1b4")
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) 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 math

torch.manual_seed(0)

# The function takes two parameters:
# 1. timesteps: a 1D tensor of timestep indices
# 2. embedding_dim: an integer specifying the embedding dimension

# Create a small 1D tensor of timesteps (batch size 2)
timesteps = torch.tensor([0, 5], dtype=torch.int64)

# Create an embedding dimension (use 4 as specified)
embedding_dim = 4

ARGS = {
    "timesteps": timesteps,
    "embedding_dim": embedding_dim,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
ismap Ran
ran, violated the contract we drafted
liyinqi/un2clip/ldm/util.py
code served (permissive licence) · get_code("d72762b700feee6f")
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 bool, 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)

# The function checks if x is a torch.Tensor with 4 dimensions where the second dimension (x.shape[1]) is > 3
# We need to construct a tensor that satisfies this condition

# Create a tensor with shape (batch=2, channels=5, height=4, width=3)
# This has 4 dimensions and x.shape[1] = 5 > 3, so it should return True
x = torch.randn(2, 5, 4, 3)

ARGS = {
    'x': x
}
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 bool, 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
l2norm Ran
ran; our contract draft was wrong, not the code
laion-ai/conditioned-prior/util.py
pointer only (licence: MIT); the stored copy get_code serves, from tmlresearchgroup-cas/clipure, is pointer only (licence: NONE) · get_code("3eb58dae620046d3")
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 tmlresearchgroup-cas/clipure, 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.

nonlinearity Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/modules/diffusionmodules/model.py
code served (permissive licence) · get_code("3137073275f8c21a")
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) 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 single parameter 'x' which is used with:
# - multiplication operator (*)
# - torch.sigmoid() function
# This means x should be a tensor

x = torch.randn(2, 4)

ARGS = {
    "x": x
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
norm_thresholding Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/models/diffusion/sampling_util.py
code served (permissive licence) · get_code("9898fe8b041f354c")
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)

# Create a small batch of tensors
# x0 should be a multi-dimensional tensor that can be flattened along dimension 1
# Shape: (batch_size=2, feature_dim=4)
x0 = torch.randn(2, 4)

# value is a scalar threshold
value = 0.5

ARGS = {
    'x0': x0,
    'value': value,
}
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
openclip_fwd Ran
ran on a synthesized input
liyinqi/un2clip/eval_mmvpvlm.py
code served (permissive licence) · get_code("1aafd47605e9ff11")
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 3, 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.nn as nn

torch.manual_seed(0)

class MockCLIPModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.projection = nn.Linear(4, 3)
    
    def forward(self, imgs, text):
        # imgs: [batch, 4]
        # text: [batch, 4]
        # Return tuple of (image_features, text_features)
        image_features = self.projection(imgs)  # [batch, 3]
        text_features = self.projection(text)   # [batch, 3]
        return (image_features, text_features)

model = MockCLIPModel()
imgs = torch.randn(2, 4)
text1 = torch.randn(2, 4)
text2 = torch.randn(2, 4)

ARGS = {
    "model": model,
    "imgs": imgs,
    "text1": text1,
    "text2": text2
}
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, ndim 3, 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
quadratic_beta_schedule Ran
ran, honoured the contract we drafted
lucidrains/DALLE2-pytorch/dalle2_pytorch/dalle2_pytorch.py
code served (permissive licence) · get_code("7bc0ce30656f4c4a")
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 (100,) 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)

ARGS = {
    "timesteps": 100
}
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 1, dtype family float, shape (100,) — shape (100,) 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
sigmoid_beta_schedule Ran
ran, honoured the contract we drafted
lucidrains/DALLE2-pytorch/dalle2_pytorch/dalle2_pytorch.py
code served (permissive licence) · get_code("48e7ff5be83206c0")
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 (10,) 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)

ARGS = {
    "timesteps": 10
}
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 1, dtype family float, shape (10,) — shape (10,) 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
slugify Ran
ran on a synthesized input
laion-ai/conditioned-prior/util.py
code served (permissive licence) · get_code("125210403ba29cf0")
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 str. 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 re
import unicodedata
from functools import lru_cache

# Define the function
@lru_cache(maxsize=None)
def slugify(value, allow_unicode=False):
    """Taken from https://github.com/django/django/blob/master/django/utils/text.py"""
    value = str(value)
    if allow_unicode:
        value = unicodedata.normalize('NFKC', value)
    else:
        value = unicodedata.normalize('NFKD', value).encode('ascii', 'ignore').decode('ascii')
    value = re.sub('[^\\w\\s-]', '', value.lower())
    return re.sub('[-\\s]+', '-', value).strip('-_')

# Create valid arguments for the function
ARGS = {
    "value": "Hello World! This is a Test-String_123",
    "allow_unicode": False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str 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
spatial_norm_thresholding Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/models/diffusion/sampling_util.py
code served (permissive licence) · get_code("d86694c13b94ae3a")
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)

# Analyze the function:
# - x0: tensor that has .pow(), .mean(), .sqrt(), .clamp(), and * operations
#   It's indexed with [1] (dimension 1) and keepdim=True
#   So x0 should be at least 2D
# - value: scalar used in clamp(min=value) and division (value / s)
#   Should be a numeric type

# Create x0 as a 2D tensor (batch_size=2, features=4)
x0 = torch.randn(2, 4)

# Create value as a small positive float
value = 0.1

ARGS = {
    "x0": x0,
    "value": value
}
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
uniform_on_device Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/models/diffusion/ddpm.py
code served (permissive licence) · get_code("d48d8354986e3b0e")
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) 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)

# Analyze the function:
# - r1: used in arithmetic (r1 - r2), should be a scalar or tensor
# - r2: used in arithmetic (r1 - r2), should be a scalar or tensor
# - shape: unpacked with *shape in torch.rand(*shape, device=device), should be a tuple/list of ints
# - device: passed to torch.rand, should be a valid torch device

ARGS = {
    "r1": 2.0,
    "r2": 1.0,
    "shape": (2, 4),
    "device": torch.device("cpu")
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
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
uniq Ran
ran; our contract draft was wrong, not the code
liyinqi/un2clip/ldm/modules/attention.py
pointer only (licence: MIT); the stored copy get_code serves, from JingWu321/MUNBa, is pointer only (licence: NONE) · get_code("9a299fe5ae09e407")
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 JingWu321/MUNBa, 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.

LongRangeModulation Not yet run
unverified (harvested, no recorded run)
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("863d6ec1313dc270")
How it ran

no invocation recorded.

ModBlock Not yet run
unverified (harvested, no recorded run)
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("ad1ef24dc899e552")
How it ran

no invocation recorded.

SteerableLRM Not yet run
unverified (harvested, no recorded run)
harvard-visionlab/lrm-steering/lrm_models/lrmnet.py
code served (permissive licence) · get_code("e85d31884a576256")
How it ran

no invocation recorded.

benchmark_model Not yet run
unverified (harvested, no recorded run)
liyinqi/un2clip/eval_mmvpvlm.py
code served (permissive licence) · get_code("730edc91bdf40519")
How it ran

no invocation recorded.

log_txt_as_img Not yet run
unverified (harvested, no recorded run)
liyinqi/un2clip/ldm/util.py
code served (permissive licence) · get_code("f9bd2e83191afad1")
How it ran

no invocation recorded.

Also present in other papers' code

Of 38 samples lifted from this paper's linked repositories, 38 carry a function identity and 0 do not. Of the 38 distinct functions with an identity, 17 are also present in other papers whose code we hold (9 with at least one paper dated earlier, 2 the same month, 17 later) and 21 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
exists
brick of this paper's assembly: helper
486 other papers
90 dated earlier, 3 the same month, 390 later, 1 same year, order not determinable, 2 undated · 78 only as an assembly brick
1603.08155 (2016-03)
Perceptual Losses for Real-Time Style Transfer and Super-Resolution
1611.07004 (2016-11)
Image-to-Image Translation with Conditional Adversarial Networks
1804.04637 (2018-04)
EMBER: An Open Dataset for Training Static PE Malware Machine Learning Models
and 483 more not named here
default 207 other papers
14 dated earlier, 193 later · 41 only as an assembly brick
1711.11585 (2017-11)
High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.08162 (2020-10)
SidechainNet: An All-Atom Protein Structure Dataset for Machine Learning
and 204 more not named here
disabled_train 173 other papers
5 dated earlier, 168 later · 5 only as an assembly brick
2012.09841 (2020-12)
Taming Transformers for High-Resolution Image Synthesis
2102.05379 (2021-02)
Argmax Flows and Multinomial Diffusion: Learning Categorical Distributions
2104.07652 (2021-04)
Geometry-Free View Synthesis: Transformers and no 3D Priors
and 170 more not named here
nonlinearity 156 other papers
10 dated earlier, 146 later · 25 only as an assembly brick
2104.07652 (2021-04)
Geometry-Free View Synthesis: Transformers and no 3D Priors
2106.00132 (2021-06)
On Fast Sampling of Diffusion Probabilistic Models
2106.06819 (2021-06)
D2C: Diffusion-Denoising Models for Few-shot Conditional Generation
and 153 more not named here
uniq 115 other papers
1 dated earlier, 114 later
2112.10752 (2021-12)
High-Resolution Image Synthesis with Latent Diffusion Models
2206.02779 (2022-06)
Blended Latent Diffusion
2208.13753 (2022-08)
Frido: Feature Pyramid Diffusion for Complex Scene Image Synthesis
and 112 more not named here
get_timestep_embedding 91 other papers
7 dated earlier, 84 later · 7 only as an assembly brick
2104.07652 (2021-04)
Geometry-Free View Synthesis: Transformers and no 3D Priors
2106.00132 (2021-06)
On Fast Sampling of Diffusion Probabilistic Models
2106.06819 (2021-06)
D2C: Diffusion-Denoising Models for Few-shot Conditional Generation
and 88 more not named here
default
brick of this paper's assembly: helper
79 other papers
4 dated earlier, 1 the same month, 74 later · 15 only as an assembly brick
1611.07004 (2016-11)
Image-to-Image Translation with Conditional Adversarial Networks
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
and 76 more not named here
ismap 69 other papers
69 later
2206.02779 (2022-06)
Blended Latent Diffusion
2208.13753 (2022-08)
Frido: Feature Pyramid Diffusion for Complex Scene Image Synthesis
2210.12965 (2022-10)
High-Resolution Image Editing via Multi-Stage Blended Diffusion
and 66 more not named here
log_txt_as_img 55 other papers
55 later
2206.02779 (2022-06)
Blended Latent Diffusion
2210.12965 (2022-10)
High-Resolution Image Editing via Multi-Stage Blended Diffusion
2212.04489 (2022-12)
SINE: SINgle Image Editing with Text-to-Image Diffusion Models
and 52 more not named here
Normalize 54 other papers
54 later · 3 only as an assembly brick
2208.13753 (2022-08)
Frido: Feature Pyramid Diffusion for Complex Scene Image Synthesis
2211.13221 (2022-11)
Latent Video Diffusion Models for High-Fidelity Long Video Generation
Xu_Versatile_Diffusion_Text_Images_and_Variations_All_in_One_Diffusion_ICCV_2023_paper (2023 (year only))
Versatile Diffusion: Text, Images and Variations All in One Diffusion Model
and 51 more not named here
uniform_on_device 44 other papers
1 dated earlier, 43 later
2112.10752 (2021-12)
High-Resolution Image Synthesis with Latent Diffusion Models
2206.02779 (2022-06)
Blended Latent Diffusion
2208.13753 (2022-08)
Frido: Feature Pyramid Diffusion for Complex Scene Image Synthesis
and 41 more not named here
append_dims 27 other papers
27 later · 2 only as an assembly brick
2304.09787 (2023-04)
NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models
2307.12493 (2023-07)
TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Composition
2308.01472 (2023-08)
Reverse Stable Diffusion: What prompt was used to generate this image?
and 24 more not named here
l2norm 16 other papers
4 dated earlier, 12 later · 4 only as an assembly brick
2112.05682 (2021-12)
Self-attention Does Not Need $O(n^2)$ Memory
2202.04200 (2022-02)
MaskGIT: Masked Generative Image Transformer
2203.07852 (2022-03)
Block-Recurrent Transformers
and 13 more not named here
autocast 14 other papers
14 later
2310.04687 (2023-10)
Targeted Attack Improves Protection against Unauthorized Diffusion Customization
2310.12190 (2023-10)
DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors
2312.00330 (2023-12)
StyleCrafter: Enhancing Stylized Text-to-Video Generation with Style Adapter
and 11 more not named here
norm_thresholding 11 other papers
11 later
2304.09787 (2023-04)
NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models
2307.12493 (2023-07)
TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Composition
2308.01472 (2023-08)
Reverse Stable Diffusion: What prompt was used to generate this image?
and 8 more not named here
spatial_norm_thresholding 9 other papers
9 later
2307.12493 (2023-07)
TF-ICON: Diffusion-Based Training-Free Cross-Domain Image Composition
2308.01472 (2023-08)
Reverse Stable Diffusion: What prompt was used to generate this image?
2312.07330 (2023-12)
Learned representation-guided diffusion models for large-image generation
and 6 more not named here
first
brick of this paper's assembly: helper
5 other papers
5 later
2303.11916 (2023-03)
CompoDiff: Versatile Composed Image Retrieval With Latent Diffusion
2308.08157 (2023-08)
Learning to Generate Semantic Layouts for Higher Text-Image Correspondence in Text-to-Image Synthesis
2310.08576 (2023-10)
Learning to Act from Actionless Videos through Dense Correspondences
and 2 more not named here

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

Contrastive models like CLIP have been shown to learn robust representations of images that capture both semantics and style. To leverage these representations for image generation, we propose a two-stage model: a prior that generates a CLIP image embedding given a text caption, and a decoder that generates an image conditioned on the image embedding. We show that explicitly generating image representations improves image diversity with minimal loss in photorealism and caption similarity. Our decoders conditioned on image representations can also produce variations of an image that preserve both its semantics and style, while varying the non-essential details absent from the image representation. Moreover, the joint embedding space of CLIP enables language-guided image manipulations in a zero-shot fashion. We use diffusion models for the decoder and experiment with both autoregressive and diffusion models for the prior, finding that the latter are computationally more efficient and produce higher-quality samples.

For agents

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

get_harvested_code_for_paper("2204.06125")
get_code_for_paper("2204.06125")
have("2204.06125")

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

curl https://syntology.ai/api/ran/2204.06125.json

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

[![Syntology run record](https://syntology.ai/api/ran/2204.06125.svg)](https://syntology.ai/paper/2204.06125)

Connect an agent — have() is free.