SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2603.22216 · 2026

Gumbel Distillation for Parallel Text Generation

Bo Liu, Qiang Liu, Chi Zhang, Xixi Hu

arXiv · PDF · Open in the Atlas

Code that ran

16 of 19 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, 1 violated it and 8 have no contract verdict; 4 failures of our instrument; 3 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/2603.22216.json · badge: /api/ran/2603.22216.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
hxixixh/gumbel-distill community
Papers with Code archive
16 ran · 3 no recorded run
FunctionStatusWhere it lives
DDiTBlock Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("86fec9b064072440")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

DDiTBlockCausal Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("8ce17998895192fd")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

DDiTFinalLayer Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("f2c2fc5cf1e7ae01")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

EmbeddingLayer Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("a26dfd38a7ca0f3f")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

GumbelNoiseEmbedder Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("f731aca918f3d54e")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

LayerNorm Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("cdb5490054e8737a")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

Rotary Ran
ran on a synthesized input · constructed an object; no result was computed
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("64d54c6f56485c6d")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

TimestepEmbedder Ran
ran on a synthesized input
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("1a545c77d3bb49b8")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

apply_rotary_pos_emb_torchscript Ran
ran, violated the contract we drafted
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("f080db2e3db70ca8")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

bias_dropout_add_scale Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("fbbc6722b74bda27")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

bias_dropout_add_scale_fused_inference Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("1db46c6c34b8dce1")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

bias_dropout_add_scale_fused_train Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("ee894c881614c3ec")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

block_diff_mask Ran
ran; our contract draft was wrong, not the code
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("14959ccf1e594504")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

modulate Ran
ran, honoured the contract we drafted
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("78b555034273f3c7")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

modulate_fused Ran
ran, honoured the contract we drafted
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("45d7695ab56ca287")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

rotate_half Ran
ran, honoured the contract we drafted
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("8526ce179df04bd4")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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.

GumbelDIT Not yet run
unverified (harvested, no recorded run)
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("776d7545fb6e9737")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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. No invocation recorded.

apply_rotary_pos_emb Not yet run
unverified (harvested, no recorded run)
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("2700f6838f384546")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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. No invocation recorded.

fused_flex_attention Not yet run
unverified (harvested, no recorded run)
hxixixh/gumbel-distill/models/dit_gumbel.py
pointer only (licence: NONE) · get_code("b4f2749f771988c2")
How it ran

The body of this sample is withheld: GitHub detected no licence in this paper's repository, hxixixh/gumbel-distill 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. No invocation recorded.

Also present in other papers' code

Of 19 samples lifted from this paper's linked repositories, 19 carry a function identity and 0 do not. Of the 19 distinct functions with an identity, 14 are also present in other papers whose code we hold (12 with at least one paper dated earlier, 1 the same month, 8 later) and 5 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
modulate
brick of this paper's assembly: helper
9 other papers
7 dated earlier, 2 later · 1 only as an assembly brick
2410.21035 (2024-10)
Beyond Autoregression: Fast LLMs via Self-Distillation Through Time
2412.10193 (2024-12)
Simple Guidance Mechanisms for Discrete Diffusion Models
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
and 6 more not named here
rotate_half
brick of this paper's assembly: helper
8 other papers
8 dated earlier · 1 only as an assembly brick
2310.16834 (2023-10)
Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
2401.12522 (2024-01)
BiTA: Bi-Directional Tuning for Lossless Acceleration in Large Language Models
2402.07754 (2024-02)
Diffusion of Thought: Chain-of-Thought Reasoning in Diffusion Language Models
and 5 more not named here
bias_dropout_add_scale
brick of this paper's assembly: helper
7 other papers
4 dated earlier, 3 later
2410.21035 (2024-10)
Beyond Autoregression: Fast LLMs via Self-Distillation Through Time
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
2505.18883 (2025-05)
Partition Generative Modeling: Masked Modeling Without Masks
and 4 more not named here
bias_dropout_add_scale_fused_train
brick of this paper's assembly: helper
5 other papers
4 dated earlier, 1 later
2410.21035 (2024-10)
Beyond Autoregression: Fast LLMs via Self-Distillation Through Time
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
2505.18883 (2025-05)
Partition Generative Modeling: Masked Modeling Without Masks
and 2 more not named here
block_diff_mask
brick of this paper's assembly: model
3 other papers
1 dated earlier, 1 the same month, 1 later
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
2603.25702 (2026-03)
S2D2: Fast Decoding for Diffusion LLMs via Training-Free Self-Speculation
2606.24982 (2026-06)
Latent Block-Diffusion Temporal Point Processes: A Semi-Autoregressive Framework for Asynchronous Event Sequence Generation
LayerNorm
brick of this paper's assembly: helper
2 other papers
1 dated earlier, 1 later · 2 only as an assembly brick
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
2606.24982 (2026-06)
Latent Block-Diffusion Temporal Point Processes: A Semi-Autoregressive Framework for Asynchronous Event Sequence Generation
bias_dropout_add_scale_fused_inference
brick of this paper's assembly: helper
2 other papers
1 dated earlier, 1 later · 1 only as an assembly brick
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
2606.24982 (2026-06)
Latent Block-Diffusion Temporal Point Processes: A Semi-Autoregressive Framework for Asynchronous Event Sequence Generation
DDiTBlock
brick of this paper's assembly: model
1 other paper
1 dated earlier · 1 only as an assembly brick
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
EmbeddingLayer
brick of this paper's assembly: helper
1 other paper
1 later · 1 only as an assembly brick
2605.31063 (2026-05)
Free energy Estimation on Any State Space
TimestepEmbedder
brick of this paper's assembly: model
1 other paper
1 later
2606.24982 (2026-06)
Latent Block-Diffusion Temporal Point Processes: A Semi-Autoregressive Framework for Asynchronous Event Sequence Generation
apply_rotary_pos_emb
brick of this paper's assembly: helper
1 other paper
1 dated earlier · 1 only as an assembly brick
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
apply_rotary_pos_emb_torchscript
brick of this paper's assembly: helper
1 other paper
1 dated earlier · 1 only as an assembly brick
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
fused_flex_attention
brick of this paper's assembly: helper
1 other paper
1 dated earlier
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS
modulate_fused
brick of this paper's assembly: helper
1 other paper
1 dated earlier · 1 only as an assembly brick
2503.09573 (2025-03)
BLOCK DIFFUSION: INTERPOLATING BETWEEN AU-TOREGRESSIVE AND DIFFUSION LANGUAGE MODELS

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

The slow, sequential nature of autoregressive (AR) language models has driven the adoption of parallel decoding methods. However, these non-AR models often sacrifice generation quality as they struggle to model the complex joint distribution of token sequences. To narrow this performance gap, we introduce Gumbel Distillation, a novel distillation technique that enables parallel decoders to learn this distribution effectively. Our method leverages the Gumbel-Max trick to create a deterministic mapping from a latent Gumbel noise space to the output tokens of a high-performing AR teacher. As a model-agnostic technique, Gumbel Distillation seamlessly integrates with diverse parallel decoding architectures, including MDLM and BD3-LM. Experiments on LM1B and OpenWebText show that Gumbel Distillation substantially improves the generation quality of parallel language models, achieving a 30.0% improvement in MAUVE score and 10.5% in generative perplexity over MDLM trained on OpenWebText dataset. Code available at: https://github.com/hxixixh/gumbel-distill. Autoregressive (AR) language models have set the standard for text generation (Radford et al., 2019;Brown et al., 2020;Touvron et al., 2023), but their sequential, token-by-token inference process introduces significant latency, hindering their use in real-world applications. To address this bottleneck, various parallel decoding methods have emerged, including Masked Diffusion Language Models (MDLMs) (

For agents

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

get_harvested_code_for_paper("2603.22216")
get_code_for_paper("2603.22216")
have("2603.22216")

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

curl https://syntology.ai/api/ran/2603.22216.json

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

[![Syntology run record](https://syntology.ai/api/ran/2603.22216.svg)](https://syntology.ai/paper/2603.22216)

Connect an agent — have() is free.