Self-Evaluation

Definitions

1. An AI system's attempt to evaluate the quality, correctness, or confidence of its own outputs. Self-evaluation can be helpful, but it is not automatically reliable and usually needs external validation too.

   In plain English: The AI system assessing the quality of its own output.

   Example: "The workflow used self-evaluation as one signal, but final acceptance still depended on schema checks and human review."

Related Terms

• AI Evaluation
• Elicitation
• Evaluation Harness
• Faithfulness
• LLM Judge
• Needle in a Haystack
• Proxy Task
• RAG Evaluation