SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2203.14206 · ICLR · 2022

Denoising Likelihood Score Matching for Conditional Score-based Data Generation

Wei-Fang Sun, Chun-Yi Lee, Chen-Hao Chao, Bo-Wun Cheng, Yi-Chen Lo, Yu-Lun Liu, Yu-Lin Chang, Chia-Ping Chen, Chia-Che Chang

arXiv · PDF · Open in the Atlas

Code that ran

2 of 2 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 0 have no contract verdict; 2 failures of our instrument; 0 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/2203.14206.json · badge: /api/ran/2203.14206.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
chen-hao-chao/dlsm canonical (classifier label)
classifier basis: classifier_2025
1 ran · 0 no recorded run
chen-hao-chao/dlsm-toy community
Papers with Code archive
1 ran · 0 no recorded run
FunctionStatusWhere it lives
get_sde_classifier_loss_fn Ran
ran; our contract draft was wrong, not the code
chen-hao-chao/dlsm/losses.py
code served (permissive licence) · get_code("751d3edb7f661ebb")
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
import torch.nn as nn
import numpy as np

torch.manual_seed(0)

# Mock mutils module since it's imported in the function
class MockMUtils:
    @staticmethod
    def get_classifier_fn(sde, classifier_model, train, continuous):
        def classifier_fn(x, t):
            return classifier_model(x, t)
        return classifier_fn
    
    @staticmethod
    def get_score_fn(sde, score_model, train, continuous):
        def score_fn(x, t):
            return score_model(x, t)
        return score_fn

import sys
from types import ModuleType
mutils = ModuleType('mutils')
mutils.get_classifier_fn = MockMUtils.get_classifier_fn
mutils.get_score_fn = MockMUtils.get_score_fn
sys.modules['mutils'] = mutils

# Mock SDE class
class MockSDE:
    def __init__(self):
        self.T = 1.0
    
    def marginal_prob(self, x, t):
        # Returns mean and std_value_cond
        batch_size = x.shape[0]
        mean = x * 0.9
        std_value_cond = torch.ones(batch_size, device=x.device) * 0.1
        return mean, std_value_cond

# Simple classifier model
class SimpleClassifier(nn.Module):
    def __init__(self, num_classes=3):
        super().__init__()
        self.fc = nn.Linear(4, num_classes)
    
    def forward(self, x, t):
        # x shape: (batch, channels, height, width) or flattened
        batch_size = x.shape[0]
        x_flat = x.view(batch_size, -1)
        return self.fc(x_flat)

# Simple score model
class SimpleScoreModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.fc = nn.Linear(4, 4)
    
    def forward(self, x, t):
        batch_size = x.shape[0]
        x_flat = x.view(batch_size, -1)
        return self.fc(x_flat)

# Create instances
sde = MockSDE()
classifier_model = SimpleClassifier(num_classes=3)
score_model = SimpleScoreModel()

# Create batch and labels
batch = torch.randn(2, 1, 2, 2)  # batch_size=2, channels=1, height=2, width=2
labels = torch.tensor([0, 1])

ARGS = {
    'sde': sde,
    'train': True,
    'num_classes': 3,
    'continuous': True,
    'loss_type': 'total',
    'weighting_dlsm': 2,
    'weighting_ce': 0,
    'coef': 0.5,
    'eps': 1e-5
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
numpy, 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_step_fn Ran
ran; our contract draft was wrong, not the code
chen-hao-chao/dlsm-toy/losses.py
code served (permissive licence) · get_code("53dc165d11a003f8")
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


class SimpleScoreModel(nn.Module):
    """Minimal score model that accepts different input signatures."""
    
    def __init__(self, input_dim):
        super().__init__()
        self.fc = nn.Linear(input_dim, input_dim)
    
    def forward(self, x, std=None):
        """Forward pass. Optionally accepts std parameter for conditioned case."""
        return self.fc(x)


torch.manual_seed(0)

# Create a minimal score model
score_model = SimpleScoreModel(input_dim=4)

# Create a batch of small training data (batch_size=2, features=4)
batch = torch.randn(2, 4)

# Create an optimizer
optimizer = torch.optim.Adam(score_model.parameters(), lr=0.001)

# Non-conditioned case arguments
ARGS = {
    "std_value": 0.1,
    "train": True,
    "conditioned": False,
    "weighting": 0
}
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

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Many existing conditional score-based data generation methods utilize Bayes' theorem to decompose the gradients of a log posterior density into a mixture of scores. These methods facilitate the training procedure of conditional score models, as a mixture of scores can be separately estimated using a score model and a classifier. However, our analysis indicates that the training objectives for the classifier in these methods may lead to a serious score mismatch issue, which corresponds to the situation that the estimated scores deviate from the true ones. Such an issue causes the samples to be misled by the deviated scores during the diffusion process, resulting in a degraded sampling quality. To resolve it, we formulate a novel training objective, called Denoising Likelihood Score Matching (DLSM) loss, for the classifier to match the gradients of the true log likelihood density. Our experimental evidence shows that the proposed method outperforms the previous methods on both Cifar-10 and Cifar-100 benchmarks noticeably in terms of several key evaluation metrics. We thus conclude that, by adopting DLSM, the conditional scores can be accurately modeled, and the effect of the score mismatch issue is alleviated.

For agents

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

get_harvested_code_for_paper("2203.14206")
get_code_for_paper("2203.14206")
have("2203.14206")

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

curl https://syntology.ai/api/ran/2203.14206.json

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

[![Syntology run record](https://syntology.ai/api/ran/2203.14206.svg)](https://syntology.ai/paper/2203.14206)

Connect an agent — have() is free.