advanced-evaluation - Advanced LLM Output Evaluation
Provides methods for scoring and comparing LLM outputs, creating rubrics, mitigating evaluation bias, and designing evaluation pipelines.
Tags
Updated: 2026-10-03Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Evaluate LLM outputs
- Compare model responses
- Create evaluation rubrics
- Mitigate evaluation bias
- Design evaluation pipelines
- Calibrate confidence scores
- Analyze judge agreement
Inputs
- Prompts and responses
- Evaluation criteria
- Scoring rubrics
- Evaluation context
Outputs
- Scores and justifications
- Pairwise verdicts
- Confidence scores
- Evaluation summaries
Requirements
- LLM judge model