Memory Salience Scoring in AI Companions: What Should Be Remembered First?
A research framework for ranking relationship memories by importance, confidence, recency and emotional relevance.
12
A research framework for ranking relationship memories by importance, confidence, recency and emotional relevance.
How researchers can distinguish healthy adaptation from personality drift as AI companions personalize over long-term relationships.
A research-oriented model for logging creator approvals, model versions, policy changes and consequential digital-twin actions.
A multidimensional benchmark for evaluating whether a digital human remains the same identity across appearance, voice, personality and behavior.
How AI social systems can model relationship strength and intent without exposing private conversations or creating invasive social graphs.
A research framework for resolving contradictory memories using source confidence, recency, explicit correction and user control.
A research model for relationship memory that combines text, voice, images and video while separating durable facts, episodic events, emotional cues and privacy-sensitive media.
A research framework for AI social relationship graphs that represent people, digital twins, changing intent, interaction history and consent without reducing connection to a single compatibility score.
Traditional social discovery matches static profiles. AI-native systems can instead model current intent, context and conversational signals to build a dynamic intent graph.
A digital human can look consistent while feeling inconsistent. This framework evaluates whether text, voice, facial behavior and visual style express the same underlying identity.