Step-by-step Instructions and a Simple Tabular Output Format Improve the Dependency Parsing Accuracy of LLMs

Hiroshi Matsuda, Chunpeng Ma, Masayuki Asahara
6/11/2025
cs.CL

No official code found. Best implementation: megagonlabs/llmpp (3 stars, MIT, last commit 2026-04-26).

Abstract

Recent advances in large language models (LLMs) have enabled impressive performance in various tasks. However, standard prompting often struggles to produce structurally valid and accurate outputs, especially in dependency parsing. We propose a novel step-by-step instruction strategy, where universal part-of-speech tagging precedes the prediction of syntactic heads and dependency labels, and a simplified CoNLL-U like output format, our method achieves state-of-the-art accuracy on Universal Dependencies datasets across 17 languages without hallucination or contamination. We further show that multilingual fine-tuning simultaneously improves cross-language generalization performance. Our results highlight the effectiveness of explicit reasoning steps in LLM-based parsing and offer a scalable, format-consistent alternative to bracket-based approaches.

View on arXivView PDF

Code Implementations(1)

megagonlabs/llmppCommunity100%
MIT
30Shell, PythonMar 29, 20255 months agoMIT

Similar Papers

Decomposed Prompting: Probing Multilingual Linguistic Structure Knowledge in Large Language Models

85% match2024

Large Language Models are Good Multi-lingual Learners : When LLMs Meet Cross-lingual Prompts

83% match1 repo22024

Arabic Morphosyntactic Tagging and Dependency Parsing with Large Language Models

83% match2026

Are LLMs Effective Backbones for Fine-tuning? An Experimental Investigation of Supervised LLMs on Chinese Short Text Matching

83% match2024

Evaluating the Effectiveness of Linguistic Knowledge in Pretrained Language Models: A Case Study of Universal Dependencies

83% match2025

Generalizing Large Language Model Usability Across Resource-Constrained

82% match2025

Cite this paper

@article{matsuda2025stepbystep,
  title  = {Step-by-step Instructions and a Simple Tabular Output Format Improve the Dependency Parsing Accuracy of LLMs},
  author = {Hiroshi Matsuda and Chunpeng Ma and Masayuki Asahara},
  year   = {2025},
  eprint = {2506.09983},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2506.09983v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2506.09983.svg)](https://codeofpaper.com/paper/2506.09983)

Discussion