AI News HubLIVE
原文2 min read

Self-Generated Error Training for Token Editing in Diffusion Language Models

This paper addresses the training-inference mismatch in the token-to-token (T2T) editor of LLaDA2.1 diffusion language models. The proposed self-generated T2T method performs a no-gradient draft pass, fills masked positions with predicted tokens, and supervises recovery in a second pass under self-generated corruptions. Implemented via LoRA continued-pretraining on LLaDA2.1-mini, it improves accuracy while reducing edit intensity, mitigating failure modes like final-digit transcription errors and excessive self-correction.

SourcearXiv Computational LinguisticsAuthor: Lin Yao

[2606.17175] Self-Generated Error Training for Token Editing in Diffusion Language Models

[Submitted on 15 Jun 2026]

Title:Self-Generated Error Training for Token Editing in Diffusion Language Models

View a PDF of the paper titled Self-Generated Error Training for Token Editing in Diffusion Language Models, by Lin Yao

View PDF HTML (experimental)

Abstract:Token-to-token (T2T) editing lets LLaDA2.1 revise committed tokens during block-diffusion decoding. The released recipe trains this editor on random vocabulary corruptions, but at inference the editor sees the model's own fluent, high-confidence draft errors instead. We study this training-inference mismatch and propose self-generated T2T, which performs a no-gradient draft pass, fills masked positions with predicted tokens, and supervises recovery in a second pass under these self-generated corruptions. We implement the update as a short LoRA continued-pretraining pass on LLaDA2.1-mini and evaluate on several benchmarks under the official Q-Mode T2T procedure with unchanged inference parameters. The method generally improves accuracy while reducing T2T edit intensity, mitigating failure modes such as final-digit transcription errors after otherwise correct reasoning and excessive self-correction before short factual answers.

Subjects:

Computation and Language (cs.CL)

Cite as: arXiv:2606.17175 [cs.CL]

(or arXiv:2606.17175v1 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2606.17175

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Lin Yao [view email] [v1] Mon, 15 Jun 2026 18:13:54 UTC (435 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Self-Generated Error Training for Token Editing in Diffusion Language Models, by Lin Yao

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.CL

new | recent | 2026-06

Change to browse by:

cs

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)