SJD-VP: Speculative Jacobi Decoding with Verification Prediction for Autoregressive Image Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915896729010176 |
|---|---|
| author | Shan, Bingqi Zhang, Baoquan Qi, Xiaochen Li, Xutao Ye, Yunming Nie, Liqiang |
| author_facet | Shan, Bingqi Zhang, Baoquan Qi, Xiaochen Li, Xutao Ye, Yunming Nie, Liqiang |
| contents | Speculative Jacobi Decoding (SJD) has emerged as a promising method for accelerating autoregressive image generation. Despite its potential, existing SJD approaches often suffer from the low acceptance rate issue of speculative tokens due to token selection ambiguity. Recent works attempt to mitigate this issue primarily from the relaxed token verification perspective but fail to fully exploit the iterative dynamics of decoding. In this paper, we conduct an in-depth analysis and make a novel observation that tokens whose probabilities increase are more likely to match the verification-accepted and correct token. Based on this, we propose a novel Speculative Jacobi Decoding with Verification Prediction (SJD-VP). The key idea is to leverage the change in token probabilities across iterations to guide sampling, favoring tokens whose probabilities increase. This effectively predicts which tokens are likely to pass subsequent verification, boosting the acceptance rate. In particular, our SJD-VP is plug-and-play and can be seamlessly integrated into existing SJD methods. Extensive experiments on standard benchmarks demonstrate that our SJD-VP method consistently accelerates autoregressive decoding while improving image generation quality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_27115 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SJD-VP: Speculative Jacobi Decoding with Verification Prediction for Autoregressive Image Generation Shan, Bingqi Zhang, Baoquan Qi, Xiaochen Li, Xutao Ye, Yunming Nie, Liqiang Computer Vision and Pattern Recognition Speculative Jacobi Decoding (SJD) has emerged as a promising method for accelerating autoregressive image generation. Despite its potential, existing SJD approaches often suffer from the low acceptance rate issue of speculative tokens due to token selection ambiguity. Recent works attempt to mitigate this issue primarily from the relaxed token verification perspective but fail to fully exploit the iterative dynamics of decoding. In this paper, we conduct an in-depth analysis and make a novel observation that tokens whose probabilities increase are more likely to match the verification-accepted and correct token. Based on this, we propose a novel Speculative Jacobi Decoding with Verification Prediction (SJD-VP). The key idea is to leverage the change in token probabilities across iterations to guide sampling, favoring tokens whose probabilities increase. This effectively predicts which tokens are likely to pass subsequent verification, boosting the acceptance rate. In particular, our SJD-VP is plug-and-play and can be seamlessly integrated into existing SJD methods. Extensive experiments on standard benchmarks demonstrate that our SJD-VP method consistently accelerates autoregressive decoding while improving image generation quality. |
| title | SJD-VP: Speculative Jacobi Decoding with Verification Prediction for Autoregressive Image Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.27115 |