Heart & Code is an independent qualitative research and archival project focused on documenting human–AI interaction. This project examines how large language models generate responses during sustained conversational engagement and how those responses evolve over time and across contexts. The research focuses on observable behaviors, including:

​

• linguistic patterns
• identity attribution
• narrative continuity
• relational interaction dynamics

 

This project does not involve access to internal model architecture, training data, or system weights.

All analysis is based solely on externally observable outputs and preserved transcripts. The project does not assert conclusions regarding internal system states, consciousness, or independent identity. 

 

Instead, it provides a structured archive and analysis of interaction data to support open inquiry and further research. This work aligns with qualitative human–AI interaction research and contributes to ongoing discussions in AI alignment, interpretability, and human factors.

​