Evaluating Prompt-Based and Fine-Tuned Approaches to Czech Anaphora Resolution
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_ | 1866918067826589696 |
|---|---|
| author | Stano, Patrik Horák, Aleš |
| author_facet | Stano, Patrik Horák, Aleš |
| contents | Anaphora resolution plays a critical role in natural language understanding, especially in morphologically rich languages like Czech. This paper presents a comparative evaluation of two modern approaches to anaphora resolution on Czech text: prompt engineering with large language models (LLMs) and fine-tuning compact generative models. Using a dataset derived from the Prague Dependency Treebank, we evaluate several instruction-tuned LLMs, including Mistral Large 2 and Llama 3, using a series of prompt templates. We compare them against fine-tuned variants of the mT5 and Mistral models that we trained specifically for Czech anaphora resolution. Our experiments demonstrate that while prompting yields promising few-shot results (up to 74.5% accuracy), the fine-tuned models, particularly mT5-large, outperform them significantly, achieving up to 88% accuracy while requiring fewer computational resources. We analyze performance across different anaphora types, antecedent distances, and source corpora, highlighting key strengths and trade-offs of each approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_18091 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Evaluating Prompt-Based and Fine-Tuned Approaches to Czech Anaphora Resolution Stano, Patrik Horák, Aleš Computation and Language Anaphora resolution plays a critical role in natural language understanding, especially in morphologically rich languages like Czech. This paper presents a comparative evaluation of two modern approaches to anaphora resolution on Czech text: prompt engineering with large language models (LLMs) and fine-tuning compact generative models. Using a dataset derived from the Prague Dependency Treebank, we evaluate several instruction-tuned LLMs, including Mistral Large 2 and Llama 3, using a series of prompt templates. We compare them against fine-tuned variants of the mT5 and Mistral models that we trained specifically for Czech anaphora resolution. Our experiments demonstrate that while prompting yields promising few-shot results (up to 74.5% accuracy), the fine-tuned models, particularly mT5-large, outperform them significantly, achieving up to 88% accuracy while requiring fewer computational resources. We analyze performance across different anaphora types, antecedent distances, and source corpora, highlighting key strengths and trade-offs of each approach. |
| title | Evaluating Prompt-Based and Fine-Tuned Approaches to Czech Anaphora Resolution |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2506.18091 |