SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · aaai_28602 · AAAI · 2024

Reducing Spatial Fitting Error in Distillation of Denoising Diffusion Models

Zhiyuan Yang, Guang Yang, Lefan Hou, Zejian Li, Lingyun Sun, Shengzhe Zhou, Shengyuan Zhang, Changyuan Yang

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 7 have no contract verdict; 3 failures of our instrument; 4 no recorded run), measured 2026-11-22; 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/aaai_28602.json · badge: /api/ran/aaai_28602.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
Sainzerjj/SFERD canonical (classifier label)
classifier basis: census
10 ran · 4 no recorded run
FunctionStatusWhere it lives
attention_from_qkv Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
Sainzerjj/SFERD/unet/teacher_unet.py
code served (permissive licence) · get_code("e2497fcb1db5da74")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 3, dtype family float, shape (4, 5, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import math
import torch as th

th.manual_seed(0)

# Construct a valid qkv tensor
# Shape: (bs, width, length) where width must be divisible by (3 * num_heads)
# Let's use: bs=2, num_heads=2, ch=4
# Then width = 3 * num_heads * ch = 3 * 2 * 4 = 24
bs = 2
num_heads = 2
ch = 4
length = 5
width = 3 * num_heads * ch

qkv = th.randn(bs, width, length)

ARGS = {
    "qkv": qkv,
    "num_heads": num_heads,
}
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 3, dtype family float, shape (4, 5, 5) — shape (4, 5, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
calc_fd Ran
ran on a synthesized input
Sainzerjj/SFERD/metrics/fid_score.py
code served (permissive licence) · get_code("932ab06d177c62c0")
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 float32, 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 numpy as np

# Set seed for reproducibility
np.random.seed(0)

# Create small valid arguments for calculate_frechet_distance / calc_fd
# The function expects:
# - mu1, mu2: 1D mean vectors (same shape)
# - sigma1, sigma2: 2D covariance matrices (same shape)
# - eps: small float for numerical stability

dim = 4

# Create mean vectors (1D arrays)
mean1 = np.random.randn(dim).astype(np.float32)
mean2 = np.random.randn(dim).astype(np.float32)

# Create positive semi-definite covariance matrices
# Generate by A @ A.T where A is random
A1 = np.random.randn(dim, dim).astype(np.float32)
sigma1 = (A1 @ A1.T).astype(np.float32)

A2 = np.random.randn(dim, dim).astype(np.float32)
sigma2 = (A2 @ A2.T).astype(np.float32)

ARGS = {
    'mean1': mean1,
    'var1': sigma1,
    'mean2': mean2,
    'var2': sigma2,
    'eps': 1e-06
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, scipy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float32, 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
calc_pr Ran
ran on a synthesized input
Sainzerjj/SFERD/metrics/precision_recall.py
code served (permissive licence) · get_code("f594a2ba23157aff")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from collections import namedtuple

torch.manual_seed(0)

Manifold = namedtuple('Manifold', ['features', 'kth'])

# Create small tensors with distinct dimensions
# manifold_1: 6 support points in 4D space
features_1 = torch.randn(6, 4)
kth_1 = torch.tensor([0.5, 0.6, 0.7, 0.8, 0.9, 1.0])

# manifold_2: 5 support points in 4D space
features_2 = torch.randn(5, 4)
kth_2 = torch.tensor([0.4, 0.5, 0.6, 0.7, 0.8])

manifold_1 = Manifold(features=features_1, kth=kth_1)
manifold_2 = Manifold(features=features_2, kth=kth_2)

ARGS = {
    'manifold_1': manifold_1,
    'manifold_2': manifold_2,
    'row_batch_size': 2,
    'col_batch_size': 2,
    'device': 'cpu'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, tqdm — 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
compute_distance Ran
ran on a synthesized input
Sainzerjj/SFERD/metrics/precision_recall.py
code served (permissive licence) · get_code("07ff17e38337d973")
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 (6, 9) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create small feature tensors
# row_features: shape (6, 4) - 6 samples, 4 features
row_features = torch.randn(6, 4)

# col_features: shape (9, 4) - 9 samples, 4 features
col_features = torch.randn(9, 4)

# Batch sizes
row_batch_size = 2
col_batch_size = 3

# Device - CPU only
device = 'cpu'

ARGS = {
    'row_features': row_features,
    'col_features': col_features,
    'row_batch_size': row_batch_size,
    'col_batch_size': col_batch_size,
    'device': device
}
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 (6, 9) — shape (6, 9) 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
data_norm Ran
ran on a synthesized input
Sainzerjj/SFERD/diffusion/utils.py
code served (permissive licence) · get_code("14693f1bdb116065")
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 tensor with both positive and negative values
x = torch.randn(2, 4)

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 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
dict2str Ran
ran on a synthesized input
Sainzerjj/SFERD/diffusion/utils.py
code served (permissive licence) · get_code("58b08cf3a865f1a0")
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 torch
import numpy as np

def dict2str(d):
    out_str = []
    for k, v in d.items():
        out_str.append(str(k))
        if isinstance(v, (list, tuple)):
            v = '_'.join(list(map(str, v)))
        elif isinstance(v, float):
            v = f'{v:.0e}'
        elif isinstance(v, dict):
            v = dict2str(v)
        out_str.append(str(v))
    out_str = '_'.join(out_str)
    return out_str


# Test with various dictionary types
ARGS = {
    'd': {
        'model_name': 'resnet50',
        'learning_rate': 0.001,
        'layers': [64, 128, 256],
        'dropout': 0.5,
        'batch_size': 32,
        'nested_config': {
            'optimizer': 'adam',
            'momentum': 0.9
        },
        'scales': (1.0, 2.0, 4.0)
    }
}
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 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
get_param Ran
ran on a synthesized input
Sainzerjj/SFERD/diffusion/utils.py
code served (permissive licence) · get_code("a80dffb347144cb4")
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 float, 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
import torch.nn as nn

torch.manual_seed(0)

# configs_1: a dict-like object that supports __getitem__
configs_1 = {
    'learning_rate': 0.001,
    'batch_size': 32,
    'epochs': 10,
    'model_type': 'transformer'
}

# configs_2: an object with attributes that support __getattribute__
class Config:
    def __init__(self):
        self.learning_rate = 0.0001
        self.batch_size = 16
        self.epochs = 5
        self.dropout = 0.1

configs_2 = Config()

# name: a parameter to look up in either configs_1 or configs_2
name = 'learning_rate'

ARGS = {
    'name': name,
    'configs_1': configs_1,
    'configs_2': configs_2
}
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 float, 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
save_input_hook Ran
ran; our contract draft was wrong, not the code
Sainzerjj/SFERD/unet/teacher_unet.py
code served (permissive licence) · get_code("25d9a8be93bc84f5")
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)

# Create a minimal module
class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 4)
    
    def forward(self, x):
        return self.linear(x)

module = SimpleModule()

# Create sample tensors for inp[0] - can be a single tensor or tuple/list
inp_tensor = torch.randn(2, 4)

# Create output tensor
out_tensor = torch.randn(2, 4)

# Package inp as a tuple (since save_input_hook accesses inp[0])
inp = (inp_tensor,)

# The out parameter
out = out_tensor

ARGS = {
    "self": module,
    "inp": inp,
    "out": out
}
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
to_uint8 Ran
ran on a synthesized input
Sainzerjj/SFERD/metrics/precision_recall.py
code served (permissive licence) · get_code("7c2211663d467709")
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 int, shape (2, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function takes a tensor x and converts it to uint8
# It applies: (x * 127.5 + 128).clamp(0, 255).to(torch.uint8)
# So x should be a float tensor, typically in range [-1, 1]

x = torch.randn(2, 4)  # Small batch of 2, dim 4

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 Tensor, ndim 2, dtype family int, 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
zero_module Ran
ran; our contract draft was wrong, not the code
Sainzerjj/SFERD/unet/student_unet.py
code served (permissive licence) · get_code("129b804760b3115f")
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 SimpleModule. 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 module with some parameters
class SimpleModule(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear = nn.Linear(4, 3)
        self.conv = nn.Conv1d(2, 2, 3)
    
    def forward(self, x):
        return x

module = SimpleModule()

# Verify the module has non-zero parameters before zeroing
for p in module.parameters():
    assert p.abs().sum() > 0, "Parameters should be non-zero initially"

ARGS = {
    "module": module
}
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 SimpleModule — a SimpleModule output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a SimpleModule 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
avg_pool_nd Not yet run
unverified (harvested, no recorded run)
Sainzerjj/SFERD/unet/student_unet.py
code served (permissive licence) · get_code("ecd0fc28815b65ae")
How it ran

no invocation recorded.

conv_nd Not yet run
unverified (harvested, no recorded run)
Sainzerjj/SFERD/unet/student_unet.py
code served (permissive licence) · get_code("fe4eb545bbb728e0")
How it ran

no invocation recorded.

get_activations Not yet run
unverified (harvested, no recorded run)
Sainzerjj/SFERD/metrics/fid_score.py
code served (permissive licence) · get_code("4c3e68855135f3e5")
How it ran

no invocation recorded.

get_precomputed Not yet run
unverified (harvested, no recorded run)
Sainzerjj/SFERD/metrics/fid_score.py
code served (permissive licence) · get_code("f352c1f6e44e6898")
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, 6 are also present in other papers whose code we hold (6 with at least one paper dated earlier, 0 the same month, 4 later) and 8 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
zero_module 196 other papers
92 dated earlier, 30 later, 74 same year, order not determinable · 49 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
2105.05233 (2021-05)
Diffusion Models Beat GANs on Image Synthesis
and 193 more not named here
conv_nd 121 other papers
66 dated earlier, 16 later, 39 same year, order not determinable · 32 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
2111.11672 (2021-11)
Few-shot Image Generation with Mixup-based Distance Learning
and 118 more not named here
avg_pool_nd 113 other papers
66 dated earlier, 12 later, 35 same year, order not determinable · 31 only as an assembly brick
2006.11239 (2020-06)
Denoising Diffusion Probabilistic Models
2010.02502 (2020-10)
Denoising Diffusion Implicit Models
2110.11291 (2021-10)
Likelihood Training of Schrödinger Bridge using Forward-Backward SDEs Theory
and 110 more not named here
get_activations 6 other papers
3 dated earlier, 2 later, 1 same year, order not determinable
Chou_How_to_Backdoor_Diffusion_Models_CVPR_2023_paper (2023 (year only))
How to Backdoor Diffusion Models?
2306.06874 (2023-06)
VillanDiffusion: A Unified Backdoor Attack Framework for Diffusion Models
2312.12030 (2023-12)
Towards Accurate Guided Diffusion Sampling through Symplectic Adjoint Method
and 3 more not named here
attention_from_qkv 2 other papers
1 dated earlier, 1 same year, order not determinable
2210.00939 (2022-10)
Improving Sample Quality of Diffusion Models Using Self-Attention Guidance
2403.17377 (2024-03)
Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance
save_input_hook 2 other papers
1 dated earlier, 1 same year, order not determinable
2210.00939 (2022-10)
Improving Sample Quality of Diffusion Models Using Self-Attention Guidance
2403.17377 (2024-03)
Self-Rectifying Diffusion Sampling with Perturbed-Attention Guidance

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

Denoising Diffusion models have exhibited remarkable capabilities in image generation. However, generating high-quality samples requires a large number of iterations. Knowledge distillation for diffusion models is an effective method to address this limitation with a shortened sampling process but causes degraded generative quality. Based on our analysis with biasvariance decomposition and experimental observations, we attribute the degradation to the spatial fitting error occurring in the training of both the teacher and student model. Accordingly, we propose Spatial Fitting-Error Reduction Distillation model (SFERD). SFERD utilizes attention guidance from the teacher model and a designed semantic gradient predictor to reduce the student's fitting error. Empirically, our proposed model facilitates high-quality sample generation in a few function evaluations. We achieve an FID of 5.31 on CIFAR-10 and 9.39 on ImageNet 64×64 with only one step, outperforming existing diffusion methods. Our study provides a new perspective on diffusion distillation by highlighting the intrinsic denoising ability of models.

For agents

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

get_harvested_code_for_paper("aaai_28602")
get_code_for_paper("aaai_28602")
have("aaai_28602")

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

curl https://syntology.ai/api/ran/aaai_28602.json

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

[![Syntology run record](https://syntology.ai/api/ran/aaai_28602.svg)](https://syntology.ai/paper/aaai_28602)

Connect an agent — have() is free.