Although auto-regressive models excel in natural language processing, they often struggle to generate diverse text and provide limited controllability. Non-auto-regressive methods could be an alternative but often produce degenerate outputs and exhibit shortcomings in conditional generation. To address these challenges, we propose Diffusion-EAGS, a novel framework that integrates conditional masked language models into diffusion language models through the theoretical lens of a conditional Markov Random Field. In doing so, we propose entropy-adaptive Gibbs sampling and entropy-based noise scheduling to counterbalance each model's shortcomings. Experimental results show that Diffusion-EAGS outperforms baselines and achieves the best quality-diversity tradeoff, demonstrating its effectiveness in non-autoregressive text generation.
@article{arxiv.2411.06438,
title = {Conditional [MASK] Discrete Diffusion Language Model},
author = {Hyukhun Koh and Minha Jhang and Dohyung Kim and Sangmook Lee and Kyomin Jung},
journal= {arXiv preprint arXiv:2411.06438},
year = {2025}
}