| title | Where Should Diffusion Enter a Language Model? Geometry-Guided Hidden-State Replacement |
|---|---|
| authors | Injin Kong, Hyoungjoon Lee, Yohan Jo |
| date | 2026-05-14 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2605.14368 |
| affiliation | HOLI-LAB |
| upvotes | 4 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Continuous diffusion language models lag behind autoregressive transformers, partly because diffusion is applied in spaces poorly suited to language denoising and token recovery. We propose DiHAL, a geometry-guided diffusion-transformer hybrid that asks where diffusion should enter a pretrained transformer. DiHAL scores layers with geometry-based proxies, selects a diffusion-friendly hidden-state interface, and replaces the lower transformer prefix with a diffusion bridge while retaining the upper layers and original LM head. By reconstructing the selected-layer hidden state rather than tokens, DiHAL avoids direct continuous-to-discrete recovery. Experiments on 8B-scale backbones show that the geometry score predicts effective shallow insertion layers under a fixed bridge-training protocol and that hidden-state recovery improves over continuous diffusion baselines in a diagnostic comparison matching the diffusion/recovery training budget. These results suggest that hidden-state geometry helps identify where diffusion-based replacement is feasible inside pretrained language models. |
cite
@misc{kong2026where,
title={Where Should Diffusion Enter a Language Model? Geometry-Guided Hidden-State Replacement},
author={Injin Kong and Hyoungjoon Lee and Yohan Jo},
year={2026},
eprint={2605.14368},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2605.14368}
}