compass-verifier-unified-llm-evaluation

Lightweight verifier model for evaluating LLM outputs across multiple domains, handling diverse answer types through meta-error pattern analysis.

Independent GitHub snapshot · Automated classification · Original documentation governs