cs.CL提交于 2026-10-31已译

面向科学家与专业人士的大语言模型入门

An Introduction to Large Language Models for Scientists and Professionals

Umberto Michelucci

中文摘要

大语言模型(LLM)已成为科学、教育、医学和专业工作中的重要工具。尽管从用户角度看似乎简单，但有效且负责任地使用需要理解这些系统如何工作、有哪些局限、输出应如何评估。本文档是作者在意大利都灵大学医学人工智能硕士课程中 LLM 讲座的基础材料，为科学家和专业人士提供实用且易读的 LLM 入门。内容解释 tokenisation、embeddings、训练、推理、上下文窗口(context window)、提示(prompting)和指令层级(instruction hierarchy)的基本概念，刻意回避数学细节。特别关注 LLM 生成答案的局限，包括幻觉(hallucination)、不确定性、可复现性、偏见，以及系统化验证和人工监督的必要性。还讨论隐私、治理、安全和敏感个人信息的负责任处理。进阶主题包括检索增强生成(RAG)、检索质量、分块(chunking)、embeddings、向量搜索、agent、工具和工作流。一个大型实践练习引导读者用 Ollama 和 Open WebUI 运行本地 LLM，并与一组科学论文交互，强调证据检索、来源验证、不确定性、研究间分歧，以及看似合理的语言与有科学依据的结论之间的区别。无需编程知识。

关键要点

01面向科学家与专业人士，用非数学方式讲解 tokenisation、embeddings、训练、推理、提示等 LLM 基础概念
02重点讨论幻觉、不确定性、可复现性、偏见等局限，强调系统验证与人工监督
03涵盖隐私、治理、安全及敏感信息处理，并介绍 RAG、向量搜索、agent 与工作流等进阶主题
04实践练习用 Ollama 和 Open WebUI 运行本地 LLM 检索科学论文，强调证据验证与研究间分歧
05全书无需编程基础，源自都灵大学医学人工智能硕士课程讲义

解读

尚无解读。

原始英文摘要

arXiv:2610.37684v1 Announce Type: cross Abstract: Large Language Models (LLMs) have rapidly become important tools for science, education, medicine, and professional work. Despite their apparent simplicity from the user's perspective, effective and responsible use requires an understanding of how these systems work, what their limitations are, and how their outputs should be evaluated. I wrote this document as a basis for my lecture on LLM in the master in artificial intelligence in medicine at the university of Turin, Italy. This book provides a practical and accessible introduction to LLMs for scientists and professionals. It explains the fundamental ideas behind tokenisation, embeddings, training, inference, context windows, prompting, and instruction hierarchies, while deliberately avoiding mathematical detail. Particular attention is given to the limitations of LLM-generated answers, including hallucinations, uncertainty, reproducibility, bias, and the need for systematic verification and human oversight. The book also discusses privacy, governance, security, and the responsible handling of personal and sensitive information. More advanced topics include Retrieval-Augmented Generation (RAG), retrieval quality, chunking, embeddings, vector search, agents, tools, and workflows. A substantial practical exercise guides the reader through running a local LLM with Ollama and Open WebUI and using it to interact with a collection of scientific papers. The exercise emphasizes evidence retrieval, source verification, uncertainty, disagreement between studies, and the distinction between plausible language and scientifically supported conclusions. No programming knowledge is required.