Across
- 3. Step-by-step thinking used to choose actions or solve problems
- 4. The ability of an AI system to choose actions toward a goal
- 5. When an agent calls APIs, search, code, or external services
- 7. Rules or checks that keep an AI system safe and on task
- 9. Measuring whether an agent performs well on a task
- 11. Framework often used to build LLM applications, chains, and agents
- 13. Component that breaks a goal into smaller steps
- 14. Stored context that helps an agent continue or personalize work
Down
- 1. Coordinating models, tools, memory, and workflows
- 2. Finding relevant information before generating an answer
- 6. Process where an agent critiques or improves its own output
- 8. Unit of text that a language model reads or generates
- 10. Open protocol that lets AI applications connect to external tools, data, and systems
- 12. Capacity to operate with limited human direction