ahmad21omar/SLR-Bench-Spanish

SLR-Bench-Spanish 是一个用于评估和训练大型语言模型（LLM）在逻辑推理方面的西班牙语数据集。它包括自然语言提示、地面真实规则、验证程序和符号等特征。数据集分为20个复杂度级别，分为四个难度级别：基础、简单、中等和困难。每个任务都附带一个验证程序用于自动评估。数据集支持自动任务生成、可编程缩放和符号自动评估。它遵循与英文版本相同的象征性结构、评估框架和课程，但所有自然语言任务提示都已翻译成西班牙语。

SLR-Bench-Spanish is a dataset designed for evaluating and training Large Language Models (LLMs) in logical reasoning using natural language task prompts. It includes features such as prompts, ground-truth rules, validation programs, and symbols, among others. The dataset is structured into a curriculum with 20 complexity levels, grouped into four tiers: basic, easy, medium, and hard. Each task is accompanied by a validation program for automatic evaluation. The dataset supports automatic task generation, programmable scaling, and symbolic automated evaluation. It follows the same symbolic structure, evaluation framework, and curriculum as the English version but provides all natural-language task prompts translated into Spanish.