Aligning Language Models for Icelandic Legal Text Summarization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Harðarson, Þórir Hrafn, Loftsson, Hrafn, Ólafsson, Stefán
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913807502147584
author Harðarson, Þórir Hrafn
Loftsson, Hrafn
Ólafsson, Stefán
author_facet Harðarson, Þórir Hrafn
Loftsson, Hrafn
Ólafsson, Stefán
contents The integration of language models in the legal domain holds considerable promise for streamlining processes and improving efficiency in managing extensive workloads. However, the specialized terminology, nuanced language, and formal style of legal texts can present substantial challenges. This study examines whether preference-based training techniques, specifically Reinforcement Learning from Human Feedback and Direct Preference Optimization, can enhance models' performance in generating Icelandic legal summaries that align with domain-specific language standards and user preferences. We compare models fine-tuned with preference training to those using conventional supervised learning. Results indicate that preference training improves the legal accuracy of generated summaries over standard fine-tuning but does not significantly enhance the overall quality of Icelandic language usage. Discrepancies between automated metrics and human evaluations further underscore the importance of qualitative assessment in developing language models for the legal domain.
format Preprint
id arxiv_https___arxiv_org_abs_2504_18180
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Aligning Language Models for Icelandic Legal Text Summarization
Harðarson, Þórir Hrafn
Loftsson, Hrafn
Ólafsson, Stefán
Computation and Language
Artificial Intelligence
Machine Learning
The integration of language models in the legal domain holds considerable promise for streamlining processes and improving efficiency in managing extensive workloads. However, the specialized terminology, nuanced language, and formal style of legal texts can present substantial challenges. This study examines whether preference-based training techniques, specifically Reinforcement Learning from Human Feedback and Direct Preference Optimization, can enhance models' performance in generating Icelandic legal summaries that align with domain-specific language standards and user preferences. We compare models fine-tuned with preference training to those using conventional supervised learning. Results indicate that preference training improves the legal accuracy of generated summaries over standard fine-tuning but does not significantly enhance the overall quality of Icelandic language usage. Discrepancies between automated metrics and human evaluations further underscore the importance of qualitative assessment in developing language models for the legal domain.
title Aligning Language Models for Icelandic Legal Text Summarization
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2504.18180