Sinn Sage -

: SaGE: Evaluating Moral Consistency in Large Language Models (March 2024).

: This research introduces Semantic Graph Entropy (SaGE) , a metric used to quantify how consistently an AI (like an LLM) responds to moral questions when they are phrased differently. sinn sage

This is the most prominent technical "long paper" currently associated with the name "SaGE." : SaGE: Evaluating Moral Consistency in Large Language

Comparing Sinn Féin between North and South - Sage Journals sinn sage