AIPrompting
-
Salience, Repetition, and Frame Adoption in Conversational AI: Threshold Failures of Interpretation
This paper examines how conversational AI behavior emerges from the interaction between prompting, memory, conversational signals, and implicit interpretive mechanisms. While prompting is commonly understood as the primary control interface, we argue that memory—particularly when shaped by high-salience signals—can significantly influence system behavior, sometimes overriding explicit user intent. Continue reading
