迈向高效、可定制且可访问的AI导师
Towards an Efficient, Customizable, and Accessible AI Tutor
AI总结：
针对LLM教育应用计算成本高的问题，提出离线RAG流水线结合小语言模型，在生物学内容上评估，并探索分块与评估改进，证明受限环境下AI导师部署可行。
摘要：
将大型语言模型（LLM）整合到教育中，为提升可访问性和参与度提供了巨大潜力，但其高昂的计算需求限制了在资源匮乏环境中的可用性，从而加剧了教育不平等。为解决这一问题，我们提出了一种离线检索增强生成（RAG）流水线，该流水线将小型语言模型（SLM）与稳健的检索机制配对，使其能够在无互联网连接的情况下提供事实准确、上下文相关的回答。我们使用特定领域的教育内容（聚焦于生物学课程）评估了该流水线的有效性。我们的分析突出了关键挑战：较小的模型（如SmolLM）难以有效利用RAG流水线提供的扩展上下文，尤其是在包含噪声或不相关块时。为提升性能，我们建议探索高级分块技术、替代的小型或量化版本的大型模型，并超越MMLU等传统指标，转向评估自由形式回答的整体评估框架。这项工作证明了在受限环境中部署AI导师的可行性，为公平、离线且基于设备的教育工具奠定了基础。
英文摘要：
The integration of large language models (LLMs) into education offers significant potential to enhance accessibility and engagement, yet their high computational demands limit usability in low-resource settings, exacerbating educational inequities. To address this, we propose an offline Retrieval-Augmented Generation (RAG) pipeline that pairs a small language model (SLM) with a robust retrieval mechanism, enabling factual, contextually relevant responses without internet connectivity. We evaluate the efficacy of this pipeline using domain-specific educational content, focusing on biology coursework. Our analysis highlights key challenges: smaller models, such as SmolLM, struggle to effectively leverage extended contexts provided by the RAG pipeline, particularly when noisy or irrelevant chunks are included. To improve performance, we propose exploring advanced chunking techniques, alternative small or quantized versions of larger models, and moving beyond traditional metrics like MMLU to a holistic evaluation framework assessing free-form response. This work demonstrates the feasibility of deploying AI tutors in constrained environments, laying the groundwork for equitable, offline, and device-based educational tools.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
