ZØN'
Independent researcher based in Taiwan. Member of IEEE, APA.
User-side contextual hallucination in AI
Examining how users experience and respond to contextual drift in conversational AI systems. Non-clinical framing focused on interaction safety and user awareness.
Conversational context frameworks
CXC-7 dimensions map the structural and semantic layers that shape how context persists and shifts within dialogue. Foundation for understanding context safety.
System-layer context evaluation
CXOD-7 and Coh(G) provide operational and coherence metrics at the system level. Enables rigorous assessment of context handling in AI architectures.
User-side protection and trust calibration
1024 Protocol and VCD offer practical defenses for maintaining boundary clarity and psychological safety. Trust calibration without clinical assumptions or pathologizing language.
Framework design and synthesis
Builds coherent models from complex systems
Systems thinker
Sees interconnections across layers and domains
Context lens
Understands attack and defense vectors
Interdisciplinary synthesis
Bridges AI safety, HCI, psychology, governance
Precise terminology
Clarity in language and conceptual boundaries
Independent execution
Conducts research without institutional constraints
