Planning-Augmented Sampling with Early Guidance for High-Reward Discovery
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866908801472397312 |
|---|---|
| author | Zhu, Rui Zhang, Yudong Yu, Xuan Zhang, Chen Wang, Xu Wang, Yang |
| author_facet | Zhu, Rui Zhang, Yudong Yu, Xuan Zhang, Chen Wang, Xu Wang, Yang |
| contents | Generative Flow Networks (GFlowNets) enable structured generation with inherent diversity, but existing sampling strategies often rely on weak guided exploration, slowing early discovery of high-reward candidates. In tasks such as molecular design, rapid and consistent generation of high-reward solutions can outweigh faithful distribution matching. We propose a planning-augmented framework in which Monte Carlo Tree Search using polynomial upper confidence bounds provides online value estimates, and a controllable soft-greedy mechanism integrates these planning signals into the GFlowNets forward policy. This design fosters early exploration of high-reward trajectories and gradually shifts to policy-driven exploitation as experience accumulates. Empirical results show that our method accelerates early high-reward discovery, sustains top-quality sample generation, and preserves diversity across representative tasks. All implementations are available at https://github.com/ZRNB/PLUS. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_00805 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Planning-Augmented Sampling with Early Guidance for High-Reward Discovery Zhu, Rui Zhang, Yudong Yu, Xuan Zhang, Chen Wang, Xu Wang, Yang Machine Learning Artificial Intelligence Generative Flow Networks (GFlowNets) enable structured generation with inherent diversity, but existing sampling strategies often rely on weak guided exploration, slowing early discovery of high-reward candidates. In tasks such as molecular design, rapid and consistent generation of high-reward solutions can outweigh faithful distribution matching. We propose a planning-augmented framework in which Monte Carlo Tree Search using polynomial upper confidence bounds provides online value estimates, and a controllable soft-greedy mechanism integrates these planning signals into the GFlowNets forward policy. This design fosters early exploration of high-reward trajectories and gradually shifts to policy-driven exploitation as experience accumulates. Empirical results show that our method accelerates early high-reward discovery, sustains top-quality sample generation, and preserves diversity across representative tasks. All implementations are available at https://github.com/ZRNB/PLUS. |
| title | Planning-Augmented Sampling with Early Guidance for High-Reward Discovery |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2510.00805 |