Modelpublic

renhouxing/ME-DLM-Stage3

sourceHugging Faceupdated 5mo agoView on Hugging Face

0likes70downloads

Model Card

Edit-Based Refinement for Parallel Masked Diffusion Language Models

<p align="center"> <a href="https://arxiv.org/abs/2605.09603">📄 Paper</a> • <a href="https://github.com/renhouxing/ME-DLM">🏠 Repo</a> • <a href="https://huggingface.co/renhouxing/ME-DLM-Stage3">🤖 Models</a> </p>

Introduction

ME-DLM is a lightweight edit-based refinement framework for masked diffusion language models. It first generates a complete response through parallel diffusion decoding, then refines the output with minimal edit operations such as replacement, deletion, and insertion, conditioned on the full sequence. By using edit distance as deterministic training supervision, ME-DLM improves sequence-level consistency while preserving the decoding efficiency of diffusion models. Built on LLaDA, it achieves consistent gains on HumanEval and GSM8K while using only one-eighth of the total diffusion steps.

Models

Model	Checkpoint
ME-DLM Stage 1	🤗 HF Link
ME-DLM Stage 2	🤗 HF Link
ME-DLM Stage 3	🤗 HF Link

Acknowledgments

We thank the following amazing projects that truly inspired us:

—LLaDA