Concept上下文学习 / In-context Learning / ICL
In-context Learning
A capability of large language models whereby task performance is substantially improved by providing a small number of examples (few-shot) within the prompt, without modifying any model weights.
Core Facts
Source Articles
Tell AI Your Intent: The Key Technique to Improve Output QualityOct 26Prompt Engineering for Absolute Beginners: Eight Chapters from Basics to MasteryOct 12Google TabFM: The "GPT-3 Moment" for Tabular Data Has ArrivedOct 11No Real Text Required: Synthetic Priors Enable Models to Learn Languages In-ContextOct 6
Claude Code高阶实战：打造顶级AI开发流Jul 22
From Prompts to Harness Engineering: The Evolution of AI Agent Development ParadigmsJul 10AI Agent三个月学习路线：从零基础到智能体开发实战Jul 10