BoundRL: Efficient Structured Text Segmentation through Reinforced Boundary Generation
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_ | 1866914478124171264 |
|---|---|
| author | Li, Haoyuan Shen, Zhengyuan Jeoung, Sullam Chen, Yueyan Li, Jiayu Zhu, Qi Wang, Shuai Ioannidis, Vassilis Rangwala, Huzefa |
| author_facet | Li, Haoyuan Shen, Zhengyuan Jeoung, Sullam Chen, Yueyan Li, Jiayu Zhu, Qi Wang, Shuai Ioannidis, Vassilis Rangwala, Huzefa |
| contents | Structured texts refer to texts containing structured elements beyond plain texts, such as code snippets and placeholders. Such structured texts increasingly require segmentation into semantically meaningful components, which cannot be effectively handled by conventional sentence-level segmentation methods. To address this, we propose BoundRL, a novel approach that jointly performs efficient token-level text segmentation and label prediction for long structured texts. Instead of generating full texts for each segment, it generates only starting tokens and reconstructs the complete texts by locating these tokens within the original texts, thereby reducing output tokens by 90% and minimizing hallucination. To train the models for the boundary generation, BoundRL~performs reinforcement learning with verifiable rewards (RLVR) that jointly optimizes document reconstruction fidelity and semantic alignment. It further mitigates entropy collapse by constructing intermediate candidates by perturbing segment boundaries and labels to create stepping stones toward higher-quality solutions. Experiments show that BoundRL enables small language models (1.7B parameters) to outperform few-shot prompting with much larger models as well as SFT and standard RLVR baselines on complex prompts used for LLM applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_20151 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | BoundRL: Efficient Structured Text Segmentation through Reinforced Boundary Generation Li, Haoyuan Shen, Zhengyuan Jeoung, Sullam Chen, Yueyan Li, Jiayu Zhu, Qi Wang, Shuai Ioannidis, Vassilis Rangwala, Huzefa Computation and Language Structured texts refer to texts containing structured elements beyond plain texts, such as code snippets and placeholders. Such structured texts increasingly require segmentation into semantically meaningful components, which cannot be effectively handled by conventional sentence-level segmentation methods. To address this, we propose BoundRL, a novel approach that jointly performs efficient token-level text segmentation and label prediction for long structured texts. Instead of generating full texts for each segment, it generates only starting tokens and reconstructs the complete texts by locating these tokens within the original texts, thereby reducing output tokens by 90% and minimizing hallucination. To train the models for the boundary generation, BoundRL~performs reinforcement learning with verifiable rewards (RLVR) that jointly optimizes document reconstruction fidelity and semantic alignment. It further mitigates entropy collapse by constructing intermediate candidates by perturbing segment boundaries and labels to create stepping stones toward higher-quality solutions. Experiments show that BoundRL enables small language models (1.7B parameters) to outperform few-shot prompting with much larger models as well as SFT and standard RLVR baselines on complex prompts used for LLM applications. |
| title | BoundRL: Efficient Structured Text Segmentation through Reinforced Boundary Generation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2510.20151 |