SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2602.22586 · 2026

TABDLM: Free-Form Tabular Data Generation via Joint Numerical-Language Diffusion

Yixin Chen, Muhan Zhang, Jiarui Feng, Yanbo Wang, Da Zheng, Donghong Cai

arXiv · PDF · Open in the Atlas

Code that ran

2 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 2 have no contract verdict; 0 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/2602.22586.json · badge: /api/ran/2602.22586.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
ilikevegetable/TabDLM community
Papers with Code archive
2 ran · 4 no recorded run
FunctionStatusWhere it lives
PowerMeanNoise Ran
ran on a synthesized input · constructed an object; no result was computed
ilikevegetable/TabDLM/tabdlm/model.py
code served (permissive licence) · get_code("1177fb5da98820fb")
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, PowerMeanNoise: 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
import abc

torch.manual_seed(0)

class Noise(abc.ABC, nn.Module):
    """
    Baseline forward method to get the total + rate of noise at a timestep
    """

    def forward(self, t):
        return (self.total_noise(t), self.rate_noise(t))

    @abc.abstractmethod
    def total_noise(self, t):
        """
        Total noise ie \\int_0^t g(t) dt + g(0)
        """
        pass

class PowerMeanNoise(Noise):
    """The noise schedule using the power mean interpolation function.
    
    This is the schedule used in EDM
    """

    def __init__(self, sigma_min=0.002, sigma_max=80, rho=7, **kwargs):
        super().__init__()
        self.sigma_min = sigma_min
        self.sigma_max = sigma_max
        self.raw_rho = rho

    def rho(self):
        return torch.tensor(self.raw_rho)

    def total_noise(self, t):
        sigma = (self.sigma_min ** (1 / self.rho()) + t * (self.sigma_max ** (1 / self.rho()) - self.sigma_min ** (1 / self.rho()))).pow(self.rho())
        return sigma

    def inverse_to_t(self, sigma):
        t = (sigma.pow(1 / self.rho()) - self.sigma_min ** (1 / self.rho())) / (self.sigma_max ** (1 / self.rho()) - self.sigma_min ** (1 / self.rho()))
        return t

ARGS = {
    "sigma_min": 0.002,
    "sigma_max": 80.0,
    "rho": 7,
}
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 PowerMeanNoise — 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 PowerMeanNoise 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
PowerMeanNoise_PerColumn Ran
ran on a synthesized input · constructed an object; no result was computed
ilikevegetable/TabDLM/tabdlm/model.py
code served (permissive licence) · get_code("2c83420a3dcb7b2f")
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, PowerMeanNoise_PerColumn: 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)

# Construct arguments for PowerMeanNoise_PerColumn
ARGS = {
    "num_numerical": 4,
    "sigma_min": 0.002,
    "sigma_max": 80,
    "rho_init": 1,
    "rho_offset": 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 PowerMeanNoise_PerColumn — 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 PowerMeanNoise_PerColumn 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
Noise Not yet run
unverified (harvested, no recorded run)
ilikevegetable/TabDLM/tabdlm/model.py
code served (permissive licence) · get_code("bcf700f63b2d2675")
How it ran

no invocation recorded.

Precond Not yet run
unverified (harvested, no recorded run)
ilikevegetable/TabDLM/tabdlm/model.py
code served (permissive licence) · get_code("62fb83b8b2a55f23")
How it ran

no invocation recorded.

TabDLM Not yet run
unverified (harvested, no recorded run)
ilikevegetable/TabDLM/tabdlm/model.py
code served (permissive licence) · get_code("d42bd5191d9f7166")
How it ran

no invocation recorded.

diffusion_model Not yet run
unverified (harvested, no recorded run)
ilikevegetable/TabDLM/tabdlm/model.py
code served (permissive licence) · get_code("3dbcc46d9871b413")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 2 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 2 the same month, 0 later) and 4 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
Noise
brick of this paper's assembly: scheduler
1 other paper
1 the same month
2602.11453 (2026-02)
From Noise to Order: Learning to Rank via Denoising Diffusion
PowerMeanNoise
brick of this paper's assembly: scheduler
1 other paper
1 the same month
2602.11453 (2026-02)
From Noise to Order: Learning to Rank via Denoising Diffusion

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

Synthetic tabular data generation has attracted growing attention due to its importance for data augmentation, foundation models, and privacy. However, real-world tabular datasets increasingly contain free-form text fields (e.g., reviews or clinical notes) alongside structured numerical and categorical attributes. Generating such heterogeneous tables with joint modeling of different modalities remains challenging. Existing approaches broadly fall into two categories: diffusion-based methods and LLM-based methods. Diffusion models can capture complex dependencies over numerical and categorical features in continuous or discrete spaces, but extending them to open-ended text is nontrivial and often leads to degraded text quality. In contrast, LLM-based generators naturally produce fluent text, yet their discrete tokenization can distort precise or wide-range numerical values, hindering accurate modeling of both numbers and language. In this work, we propose TABDLM, a unified framework for free-form tabular data generation via a joint numerical-language diffusion model built on masked diffusion language models (MDLMs). TABDLM models textual and categorical features through masked diffusion, while modeling numerical features with a continuous diffusion process through learned specialized numeric tokens embedding; bidirectional attention then captures cross-modality interactions within a single model. Extensive experiments on diverse benchmarks demonstrate the effectiveness of TABDLM compared to strong diffusion-and LLM-based baselines. The code is available at https://github.com/ilikevegetable/TabDLM

For agents

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

get_harvested_code_for_paper("2602.22586")
get_code_for_paper("2602.22586")
have("2602.22586")

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

curl https://syntology.ai/api/ran/2602.22586.json

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

[![Syntology run record](https://syntology.ai/api/ran/2602.22586.svg)](https://syntology.ai/paper/2602.22586)

Connect an agent — have() is free.