Automata-Based Steering of Large Language Models for Diverse Structured Generation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918202001326080 |
|---|---|
| author | Luan, Xiaokun Wei, Zeming Zhang, Yihao Sun, Meng |
| author_facet | Luan, Xiaokun Wei, Zeming Zhang, Yihao Sun, Meng |
| contents | Large language models (LLMs) are increasingly tasked with generating structured outputs. While structured generation methods ensure validity, they often lack output diversity, a critical limitation that we confirm in our preliminary study. We propose a novel method to enhance diversity in automaton-based structured generation. Our approach utilizes automata traversal history to steer LLMs towards novel structural patterns. Evaluations show our method significantly improves structural and content diversity while maintaining comparable generation efficiency. Furthermore, we conduct a case study showcasing the effectiveness of our method in generating diverse test cases for testing open-source libraries. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_11018 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Automata-Based Steering of Large Language Models for Diverse Structured Generation Luan, Xiaokun Wei, Zeming Zhang, Yihao Sun, Meng Computation and Language Artificial Intelligence Cryptography and Security Machine Learning Software Engineering Large language models (LLMs) are increasingly tasked with generating structured outputs. While structured generation methods ensure validity, they often lack output diversity, a critical limitation that we confirm in our preliminary study. We propose a novel method to enhance diversity in automaton-based structured generation. Our approach utilizes automata traversal history to steer LLMs towards novel structural patterns. Evaluations show our method significantly improves structural and content diversity while maintaining comparable generation efficiency. Furthermore, we conduct a case study showcasing the effectiveness of our method in generating diverse test cases for testing open-source libraries. |
| title | Automata-Based Steering of Large Language Models for Diverse Structured Generation |
| topic | Computation and Language Artificial Intelligence Cryptography and Security Machine Learning Software Engineering |
| url | https://arxiv.org/abs/2511.11018 |