SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2303.08133 · ICLR · 2023

MeshDiffusion: Score-based Generative 3D Mesh Modeling

arXiv · PDF · Open in the Atlas

Code that ran

10 of 14 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 3 have no contract verdict; 7 failures of our instrument; 4 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/2303.08133.json · badge: /api/ran/2303.08133.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
lzzcd001/MeshDiffusion canonical (classifier label)
classifier basis not recorded on the link
8 ran · 1 no recorded run
GuyTevet/motion-diffusion-model canonical (classifier label)
classifier basis: classifier_2025
1 ran · 3 no recorded run
copy not recorded community
not in the paper's declared repositories
1 ran · 0 no recorded run
FunctionStatusWhere it lives
construct_template_variables Ran
ran; our contract draft was wrong, not the code
GuyTevet/motion-diffusion-model/sample/generate.py
code served (permissive licence) · get_code("fe28ada12e5b4764")
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 tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function takes a single boolean parameter 'unconstrained'
# It constructs and returns template strings based on this parameter
# No complex objects are needed - just a boolean value

ARGS = {
    'unconstrained': 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 tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_ddpm_loss_fn Ran
ran on a synthesized input
lzzcd001/MeshDiffusion/lib/diffusion/losses.py
code served (permissive licence) · get_code("765e2b314d64f9a6")
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 numpy as np
import torch.nn as nn

torch.manual_seed(0)

# Create a minimal mock for mutils.get_model_fn
class mutils:
    @staticmethod
    def get_model_fn(model, train=False):
        def model_fn(x, t):
            return model(x, t)
        return model_fn

# Create a minimal VPSDE mock
class MockVPSDE:
    def __init__(self):
        self.N = 8  # number of timesteps
        # Create sqrt_alphas_cumprod and sqrt_1m_alphas_cumprod
        # Shape should be (N,) to allow indexing with labels
        self.sqrt_alphas_cumprod = torch.linspace(0.9, 0.1, self.N)
        self.sqrt_1m_alphas_cumprod = torch.linspace(0.1, 0.9, self.N)

# Create a minimal model that takes perturbed_data and labels
class MockDiffusionModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
    
    def forward(self, x, t):
        # x shape: (batch, 1, 1, 1, 4) after the unsqueezes in perturbed_data
        # Reshape to (batch, 4) for linear layer
        batch_size = x.shape[0]
        x_flat = x.reshape(batch_size, -1)
        out = self.linear(x_flat)
        # Reshape back to input shape
        return out.reshape(x.shape)

# Create test batch with shape (batch_size, channels, height, width, features)
# Using batch=2, and distinct dimension sizes for probing
batch_data = torch.randn(2, 1, 1, 1, 4)

# Create mask (optional, but used in the function)
mask = torch.ones(2, 1, 1, 1, 4)

# Create VPSDE instance
vpsde = MockVPSDE()

# Create model instance
model = MockDiffusionModel()

# Define arguments dictionary
ARGS = {
    'vpsde': vpsde,
    'train': True,
    'mask': mask,
    'loss_type': 'l2'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type 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
get_div_fn Ran
ran; our contract draft was wrong, not the code
lzzcd001/MeshDiffusion/lib/diffusion/likelihood.py
pointer only (licence: MIT); the stored copy get_code serves, from rarazafin/score_diffusion_ensemble, is pointer only (licence: NONE) · get_code("0bfcb6b2d940e393")
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 rarazafin/score_diffusion_ensemble, 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.

get_likelihood_fn Ran
ran on a synthesized input
lzzcd001/MeshDiffusion/lib/diffusion/likelihood.py
code served (permissive licence) · get_code("d613d83520b430a7")
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
import numpy as np

torch.manual_seed(0)

# Create a minimal SDE class that has the required interface
class MinimalSDE:
    def __init__(self):
        self.T = 1.0
    
    def reverse(self, score_fn, probability_flow=True):
        return self
    
    def sde(self, x, t):
        # Returns (drift, diffusion)
        drift = torch.zeros_like(x)
        diffusion = torch.ones(x.shape[0], 1)
        return drift, diffusion
    
    def prior_logp(self, z):
        # Standard normal prior
        return -0.5 * torch.sum(z ** 2, dim=tuple(range(1, len(z.shape))))

# Create a minimal score model
class MinimalScoreModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
    
    def forward(self, x, t):
        return self.linear(x)

# Create a minimal mutils module that the function uses
class mutils:
    @staticmethod
    def get_score_fn(sde, model, train=False, continuous=True):
        def score_fn(x, t):
            return model(x, t)
        return score_fn
    
    @staticmethod
    def to_flattened_numpy(x):
        if isinstance(x, torch.Tensor):
            return x.detach().cpu().numpy().reshape(x.shape[0], -1)
        return x.reshape(x.shape[0], -1)
    
    @staticmethod
    def from_flattened_numpy(x, shape):
        return torch.from_numpy(x).reshape(shape)

# Inject mutils into the module namespace where get_likelihood_fn expects it
import sys
current_module = sys.modules[__name__]
current_module.mutils = mutils

# Create test data
data = torch.randn(2, 4)

# Create SDE and model
sde = MinimalSDE()
model = MinimalScoreModel()

# Define inverse scaler
def inverse_scaler(x):
    return x * 0.5 + 1.0

# Build ARGS dict with all parameters
ARGS = {
    'sde': sde,
    'inverse_scaler': inverse_scaler,
    'hutchinson_type': 'Rademacher',
    'rtol': 1e-05,
    'atol': 1e-05,
    'method': 'RK45',
    'eps': 1e-05,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy, 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
get_optimizer Ran
ran; our contract draft was wrong, not the code
lzzcd001/MeshDiffusion/lib/diffusion/losses.py
pointer only (licence: MIT); the stored copy get_code serves, from hfmei/idc, is pointer only (licence: NONE) · get_code("8ff7c24dade9904c")
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 hfmei/idc, 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.

optimization_manager Ran
ran; our contract draft was wrong, not the code
lzzcd001/MeshDiffusion/lib/diffusion/losses.py
pointer only (licence: MIT); the stored copy get_code serves, from hfmei/idc, is pointer only (licence: NONE) · get_code("366bb8c7035a2fc8")
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 hfmei/idc, 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.

parse_comma_separated_list Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from nv-tlabs/GET3D
pointer only; the stored copy get_code serves, from nv-tlabs/GET3D, is pointer only (licence: NOASSERTION) · get_code("d01b68a634f71551")
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.

register_corrector Ran
ran; our contract draft was wrong, not the code
lzzcd001/MeshDiffusion/lib/diffusion/sampling.py
code served (permissive licence) · get_code("a05b1d0cf3f508b4")
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 type. 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)

# Clear the registry for clean test
_CORRECTORS = {}

def register_corrector(cls=None, *, name=None):
    """A decorator for registering corrector classes."""

    def _register(cls):
        if name is None:
            local_name = cls.__name__
        else:
            local_name = name
        if local_name in _CORRECTORS:
            raise ValueError(f'Already registered model with name: {local_name}')
        _CORRECTORS[local_name] = cls
        return cls
    if cls is None:
        return _register
    else:
        return _register(cls)


# Create a minimal corrector class for testing
class SimpleCorrector:
    """A minimal corrector class."""
    pass


# Test case 1: Using decorator with cls directly (no name parameter)
ARGS = {
    'cls': SimpleCorrector,
    'name': None
}
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 type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
register_predictor Ran
ran; our contract draft was wrong, not the code
lzzcd001/MeshDiffusion/lib/diffusion/sampling.py
code served (permissive licence) · get_code("473066ddf62a075a")
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 type. 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)

# Clear the registry for clean test
_PREDICTORS = {}

def register_predictor(cls=None, *, name=None):
    """A decorator for registering predictor classes."""

    def _register(cls):
        if name is None:
            local_name = cls.__name__
        else:
            local_name = name
        if local_name in _PREDICTORS:
            raise ValueError(f'Already registered model with name: {local_name}')
        _PREDICTORS[local_name] = cls
        return cls
    if cls is None:
        return _register
    else:
        return _register(cls)


# Create a simple predictor class to register
class SimplePredictor(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 2)
    
    def forward(self, x):
        return self.linear(x)


# Test case 1: Using as decorator without arguments (cls is passed directly)
# This tests the else branch where cls is not None
ARGS = {
    "cls": SimplePredictor,
    "name": None
}
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 type — a type output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a type 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
slerp Ran
ran on a synthesized input
lzzcd001/MeshDiffusion/lib/diffusion/evaler.py
code served (permissive licence) · get_code("445022320acba4c2")
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 (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 two random latent vectors of the same dimension
# The function performs slerp between z1 and z2
z1 = torch.randn(4)
z2 = torch.randn(4)

# Normalize them to avoid numerical issues in the slerp computation
z1 = z1 / torch.norm(z1)
z2 = z2 / torch.norm(z2)

# Alpha is the interpolation parameter, typically in [0, 1]
alpha = 0.5

ARGS = {
    "z1": z1,
    "z2": z2,
    "alpha": alpha,
}
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 (4,) — shape (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
evaluate_diversity Not yet run
unverified (harvested, no recorded run)
GuyTevet/motion-diffusion-model/eval/eval_humanml.py
code served (permissive licence) · get_code("0eba32186c812b63")
How it ran

no invocation recorded.

evaluate_fid Not yet run
unverified (harvested, no recorded run)
GuyTevet/motion-diffusion-model/eval/eval_humanml.py
code served (permissive licence) · get_code("6a6a115008da7ad3")
How it ran

no invocation recorded.

evaluate_matching_score Not yet run
unverified (harvested, no recorded run)
GuyTevet/motion-diffusion-model/eval/eval_humanml.py
code served (permissive licence) · get_code("0f0a2de622a554fe")
How it ran

no invocation recorded.

get_predictor Not yet run
unverified (harvested, no recorded run)
lzzcd001/MeshDiffusion/lib/diffusion/sampling.py
code served (permissive licence) · get_code("a58facb8fe71823a")
How it ran

no invocation recorded.

Also present in other papers' code

Of 14 samples lifted from this paper's linked repositories, 14 carry a function identity and 0 do not. Of the 14 distinct functions with an identity, 11 are also present in other papers whose code we hold (8 with at least one paper dated earlier, 4 the same month, 11 later) and 3 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
get_predictor 16 other papers
7 dated earlier, 1 the same month, 8 later · 1 only as an assembly brick
2006.09011 (2020-06)
Improved Techniques for Training Score-Based Generative Models
2105.14080 (2021-05)
Gotta Go Fast When Generating Data with Score-Based Models
2112.07068 (2021-12)
Score-Based Generative Modeling with Critically-Damped Langevin Diffusion
and 13 more not named here
optimization_manager 16 other papers
8 dated earlier, 7 later, 1 same year, order not determinable
2106.05527 (2021-06)
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation
2110.05243 (2021-10)
Score-based diffusion models for accelerated MRI
2205.01490 (2022-05)
Subspace Diffusion Generative Models
and 13 more not named here
get_optimizer 15 other papers
6 dated earlier, 8 later, 1 same year, order not determinable
2110.05243 (2021-10)
Score-based diffusion models for accelerated MRI
2205.01490 (2022-05)
Subspace Diffusion Generative Models
2209.12753 (2022-09)
On Investigating the Conservative Property of Score-Based Generative Models
and 12 more not named here
register_corrector 15 other papers
7 dated earlier, 1 the same month, 7 later · 2 only as an assembly brick
2006.09011 (2020-06)
Improved Techniques for Training Score-Based Generative Models
2011.13456 (2020-11)
Score-Based Generative Modeling through Stochastic Differential Equations
2105.14080 (2021-05)
Gotta Go Fast When Generating Data with Score-Based Models
and 12 more not named here
register_predictor 15 other papers
7 dated earlier, 1 the same month, 7 later · 2 only as an assembly brick
2006.09011 (2020-06)
Improved Techniques for Training Score-Based Generative Models
2011.13456 (2020-11)
Score-Based Generative Modeling through Stochastic Differential Equations
2105.14080 (2021-05)
Gotta Go Fast When Generating Data with Score-Based Models
and 12 more not named here
parse_comma_separated_list 13 other papers
6 dated earlier, 7 later
2112.07945 (2021-12)
Efficient Geometry-aware 3D Generative Adversarial Networks
2202.00273 (2022-02)
StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets
2206.07695 (2022-06)
VoxGRAF: Fast 3D-Aware Image Synthesis with Sparse Voxel Grids
and 10 more not named here
get_div_fn 11 other papers
6 dated earlier, 4 later, 1 same year, order not determinable
2106.05527 (2021-06)
Soft Truncation: A Universal Training Technique of Score-based Diffusion Model for High Precision Score Estimation
2205.01490 (2022-05)
Subspace Diffusion Generative Models
2205.13699 (2022-05)
Maximum Likelihood Training of Implicit Nonlinear Diffusion Models
and 8 more not named here
construct_template_variables 5 other papers
2 dated earlier, 3 later
2209.14916 (2022-09)
Human Motion Diffusion Model
2302.05905 (2023-02)
Single Motion Diffusion
2308.01850 (2023-08)
Synthesizing Long-Term Human Motions with Diffusion Models via Coherent Sampling
and 2 more not named here
evaluate_diversity 3 other papers
1 the same month, 2 later
2303.01418 (2023-03)
Human Motion Diffusion as a Generative Prior
2308.01850 (2023-08)
Synthesizing Long-Term Human Motions with Diffusion Models via Coherent Sampling
2312.02256 (2023-12)
EMDM: Efficient Motion Diffusion Model for Fast and High-Quality Motion Generation
evaluate_fid 1 other paper
1 later
2312.02256 (2023-12)
EMDM: Efficient Motion Diffusion Model for Fast and High-Quality Motion Generation
evaluate_matching_score 1 other paper
1 later
2312.02256 (2023-12)
EMDM: Efficient Motion Diffusion Model for Fast and High-Quality Motion Generation

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

We consider the task of generating realistic 3D shapes, which is useful for a variety of applications such as automatic scene generation and physical simulation. Compared to other 3D representations like voxels and point clouds, meshes are more desirable in practice, because (1) they enable easy and arbitrary manipulation of shapes for relighting and simulation, and (2) they can fully leverage the power of modern graphics pipelines which are mostly optimized for meshes. Previous scalable methods for generating meshes typically rely on sub-optimal post-processing, and they tend to produce overly-smooth or noisy surfaces without fine-grained geometric details. To overcome these shortcomings, we take advantage of the graph structure of meshes and use a simple yet very effective generative modeling method to generate 3D meshes. Specifically, we represent meshes with deformable tetrahedral grids, and then train a diffusion model on this direct parametrization. We demonstrate the effectiveness of our model on multiple generative tasks.

For agents

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

get_harvested_code_for_paper("2303.08133")
get_code_for_paper("2303.08133")
have("2303.08133")

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

curl https://syntology.ai/api/ran/2303.08133.json

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

[![Syntology run record](https://syntology.ai/api/ran/2303.08133.svg)](https://syntology.ai/paper/2303.08133)

Connect an agent — have() is free.