IAAR-Shanghai/VAR

xVerify 是一个从预训练大型语言模型微调的评价工具，专为具有单个正确答案的客观问题设计。它能够高效处理长推理链，从复杂的推理过程中提取最终答案，并支持多种语言，准确识别等价表达式的不同表示。此外，它还提供多种模型变体，具有不同的架构和大小，以减轻基础模型偏差。VAR 数据集用于 xVerify 的训练、测试和泛化，包含高质量的 LLM 推理和响应验证示例，涵盖各种类型的客观问题。

xVerify is an evaluation tool fine-tuned from a pre-trained large language model, designed specifically for objective questions with a single correct answer. It accurately extracts the final answer from lengthy reasoning processes and efficiently identifies equivalence across different forms of mathematical expressions, LaTeX and string representations, as well as natural language descriptions through intelligent equivalence comparison. Whether its mathematical problems, multiple-choice questions, classification tasks, or short-answer questions, xVerify delivers precise and flexible answer evaluation, significantly enhancing assessment efficiency and accuracy.