Installs as a private draft. Your edits and self-improvement runs do not change the published bundle.
Use a language model to score agent outputs against specifications or rubrics
Use a language model to score agent outputs against specifications or rubrics
When programmatic scoring is impossible, LLM judges fill the gap
Reward Engineering > Reward Function Design
--- name: "Llm As Judge Reward" version: "1.0.0" description: "Use a language model to score agent outputs against specifications or rubrics" tags: - rl-skills-universe - "reward-engineering" - "reward-function-design" trigger: "llm-as-judge-reward" --- # Llm As Judge Reward Use a language model to score agent outputs against specifications or rubrics ## Why It Matters When programmatic scoring is impossible, LLM judges fill the gap ## Domain **Reward Engineering** > Reward Function Design