CardiffNLP at CLEARS-2025: Prompting Large Language Models for Plain Language and Easy-to-Read Text Rewriting

TL;DR

This paper describes CardiffNLP's approach to using prompting techniques with large language models to rewrite Spanish texts into plain and easy-to-read language for the CLEARS-2025 shared task, achieving top placements.

Contribution

The paper introduces prompt-based methods with various prompt variations for large language models to perform Spanish text simplification, demonstrating competitive results in shared task settings.

Findings

Achieved third place in Subtask 1 and second in Subtask 2.

Explored multiple prompt variations and experimental setups.

Used Gemma-3 model for final submissions.

Abstract

This paper details the CardiffNLP team's contribution to the CLEARS shared task on Spanish text adaptation, hosted by IberLEF 2025. The shared task contained two subtasks and the team submitted to both. Our team took an LLM-prompting approach with different prompt variations. While we initially experimented with LLaMA-3.2, we adopted Gemma-3 for our final submission, and landed third place in Subtask 1 and second place in Subtask 2. We detail our numerous prompt variations, examples, and experimental results.