advanced-evaluation

SKILL.md

LLM-as-a-Judge techniques for evaluating AI outputs. Not a single technique but a family of approaches - choosing the right one and mitigating biases is the core competency.

Direct Scoring: Single LLM rates one response on a defined scale.

Pairwise Comparison: LLM compares two responses and selects better one.

掌握法學碩士法官評估技術，包括直接評分、成對比較、評分細則生成和偏差緩解。在構建評估系統、比較模型輸出或為人工智能生成的內容建立質量標準時使用。 來源：shipshitdev/library。

查看原文