Codifying Natural Langauge Tasks
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866909799508082688 |
|---|---|
| author | Chen, Haoyang Tanaka-Ishii, Kumiko |
| author_facet | Chen, Haoyang Tanaka-Ishii, Kumiko |
| contents | We explore the applicability of text-to-code to solve real-world problems that are typically solved in natural language, such as legal judgment and medical QA. Unlike previous works, our approach leverages the explicit reasoning provided by program generation. We present ICRAG, a framework that transforms natural language into executable programs through iterative refinement using external knowledge from domain resources and GitHub. Across 13 benchmarks, ICRAG achieves up to 161.1\% relative improvement. We provide a detailed analysis of the generated code and the impact of external knowledge, and we discuss the limitations of applying text-to-code approaches to real-world natural language tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_17455 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Codifying Natural Langauge Tasks Chen, Haoyang Tanaka-Ishii, Kumiko Computation and Language Artificial Intelligence We explore the applicability of text-to-code to solve real-world problems that are typically solved in natural language, such as legal judgment and medical QA. Unlike previous works, our approach leverages the explicit reasoning provided by program generation. We present ICRAG, a framework that transforms natural language into executable programs through iterative refinement using external knowledge from domain resources and GitHub. Across 13 benchmarks, ICRAG achieves up to 161.1\% relative improvement. We provide a detailed analysis of the generated code and the impact of external knowledge, and we discuss the limitations of applying text-to-code approaches to real-world natural language tasks. |
| title | Codifying Natural Langauge Tasks |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2509.17455 |