| title | Triplet-Block Diffusion RWKV |
|---|---|
| authors | Ke Lin, Yiyang Luo, Zhaolong Su, Yunya Song, Anyi Rao |
| date | 2026-05-25 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2605.25969 |
| upvotes | 11 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf code |
| abstract | Causal Transformer language models suffer from strictly sequential decoding and a quadratic per-step attention cost. While linear-time causal models and discrete diffusion models each address these weaknesses, their integration remains inherently inconsistent: diffusion requires bidirectional attention, while causal models are unidirectional. To unify these architectures, we propose B^3D-RWKV, a diffusion RWKV variant that integrates the model's O(L) inference efficiency with parallel, bidirectional discrete-diffusion through a triplet-block layout method. B^3D-RWKV-7.2B reaches comparable accuracy on an 8-task suite versus existing models while significantly outperforming baselines in decoding throughput with an average of 1.6times speedup. |
cite
@misc{lin2026triplet,
title={Triplet-Block Diffusion RWKV},
author={Ke Lin and Yiyang Luo and Zhaolong Su and Yunya Song and Anyi Rao},
year={2026},
eprint={2605.25969},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2605.25969}
}