目的评估主流大语言模型(large language model,LLM)在医学检验教育中的专业知识水平,探究其是否具备辅导医学检验技术专业学习的可能性。方法选取2023年我国临床医学检验技术(师)资格考试真题共400题,分2轮测试5款LLM(Copilot、Grok、...目的评估主流大语言模型(large language model,LLM)在医学检验教育中的专业知识水平,探究其是否具备辅导医学检验技术专业学习的可能性。方法选取2023年我国临床医学检验技术(师)资格考试真题共400题,分2轮测试5款LLM(Copilot、Grok、元宝、豆包及Kimi):首轮采用零提示词策略,次轮采用交互优化策略,评估其答题正确率及生成内容质量。通过Cochran’sQ检验比较LLM性能差异,并基于CLEAR框架(完整性、准确性、循证性、恰当性、相关性)对生成内容进行评分。结果首轮测试中,豆包总正确率最高(375/400),且豆包与元宝的总正确率显著优于Copilot与Kimi(P<0.001)。次轮优化交互后,Kimi的正确率显著提升(P<0.05);其他LLM的正确率均略有提升,但差异无统计学意义(P>0.05)。豆包的总正确率仍最高(380/400),豆包和元宝的总正确率显著高于Copilot(P<0.005)。基于CLEAR框架的评估显示,元宝、豆包、Kimi在循证性(P<0.003)和完整性(P<0.05)维度显著优于国外LLM,能规范引用权威证据,生成内容质量更优。结论主流LLM具备丰富的医学检验知识,通过单题输入、明确证据要求及启用高级推理功能可提升答题正确率与生成内容质量。国产LLM在正确率上与国外LLM相当,且在循证性和完整性维度具有显著优势。研究表明,主流LLM可作为辅助工具用于医学检验技术专业知识学习。展开更多
The teaching of grammar has received increasing attention in recent language teaching and learning literature. The major argument lies in teaching grammar as product or as process. Teaching grammar as product focuses ...The teaching of grammar has received increasing attention in recent language teaching and learning literature. The major argument lies in teaching grammar as product or as process. Teaching grammar as product focuses on giving learners a clear and explicit framework about the language; while teaching grammar as a process emphasizes the use of language by the learner. This paper gives a brief introduction of teaching grammar as product and as process and points out that language teachers can choose the more appropriate approach of them in the specific context.展开更多
文摘目的评估主流大语言模型(large language model,LLM)在医学检验教育中的专业知识水平,探究其是否具备辅导医学检验技术专业学习的可能性。方法选取2023年我国临床医学检验技术(师)资格考试真题共400题,分2轮测试5款LLM(Copilot、Grok、元宝、豆包及Kimi):首轮采用零提示词策略,次轮采用交互优化策略,评估其答题正确率及生成内容质量。通过Cochran’sQ检验比较LLM性能差异,并基于CLEAR框架(完整性、准确性、循证性、恰当性、相关性)对生成内容进行评分。结果首轮测试中,豆包总正确率最高(375/400),且豆包与元宝的总正确率显著优于Copilot与Kimi(P<0.001)。次轮优化交互后,Kimi的正确率显著提升(P<0.05);其他LLM的正确率均略有提升,但差异无统计学意义(P>0.05)。豆包的总正确率仍最高(380/400),豆包和元宝的总正确率显著高于Copilot(P<0.005)。基于CLEAR框架的评估显示,元宝、豆包、Kimi在循证性(P<0.003)和完整性(P<0.05)维度显著优于国外LLM,能规范引用权威证据,生成内容质量更优。结论主流LLM具备丰富的医学检验知识,通过单题输入、明确证据要求及启用高级推理功能可提升答题正确率与生成内容质量。国产LLM在正确率上与国外LLM相当,且在循证性和完整性维度具有显著优势。研究表明,主流LLM可作为辅助工具用于医学检验技术专业知识学习。
文摘The teaching of grammar has received increasing attention in recent language teaching and learning literature. The major argument lies in teaching grammar as product or as process. Teaching grammar as product focuses on giving learners a clear and explicit framework about the language; while teaching grammar as a process emphasizes the use of language by the learner. This paper gives a brief introduction of teaching grammar as product and as process and points out that language teachers can choose the more appropriate approach of them in the specific context.
