Taxonomy of behaviors
A system for classifying observed actions or characteristics in human-AI interactions.
A taxonomy of behaviors is like a detailed list that helps people understand and categorize the different ways AI systems act or how people interact with them. Think of it as a way to sort and label specific actions, like an AI generating text, answering a question, or even making a mistake.
For example, if you're using an AI chatbot, a taxonomy of behaviors might include categories like 'information retrieval,' 'creative writing,' 'problem solving,' or 'error handling.' This helps researchers and developers analyze how the AI is performing and how users are engaging with it.
A taxonomy of behaviors provides a structured framework for classifying and analyzing the observable actions and interaction patterns between humans and AI systems. It serves as a common language for researchers and practitioners to describe, measure, and compare AI performance and user engagement across different tasks and contexts.
These taxonomies are crucial for systematic evaluation, identifying emergent capabilities, and diagnosing failure modes. They can range from broad categories of AI actions to granular descriptions of user prompts and AI responses, enabling detailed analysis of interaction dynamics.
Anthropic's AI Fluency Index uses a taxonomy of behaviors to understand how users interact with Claude. This focus on user behavior and skill development is becoming more important as AI tools become more integrated into daily workflows and education.
- Anthropic Education Report: The AI Fluency Index · Claude AcademyAnthropic · Anthropic Research ·
- Anthropic Claude
- Anthropic uses a taxonomy of behaviors to assess and improve user interaction with Claude, as seen in their AI Fluency Index.
The concept of taxonomies of behaviors originates from fields like ethology (the study of animal behavior) and psychology, adapted for the study of human-computer interaction and now AI interaction.
Ask an AI assistant to perform a specific task, like writing a poem or explaining a complex concept. Then, try to categorize the AI's response based on its apparent function (e.g., creative generation, factual explanation, summarization). Repeat with different types of requests to observe varied behaviors.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is an example of a behavior in an AI taxonomy?
- An example behavior could be 'providing a step-by-step solution to a math problem,' 'generating a story based on a prompt,' or 'refusing to answer an inappropriate request.'
- Why are taxonomies of behaviors important for AI development?
- They help developers understand how AI systems perform, identify areas for improvement, and ensure the AI behaves in desired ways, especially in complex or safety-critical applications.
First seen on Subvolts Sep 11, 2026. Updated Sep 18, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.