감성공명 ON/OFF 트리거에 반응하는 AGI의 자율상태 전이 메카니즘

초록

This study presents an empirical exploration of the JRESA architecture to address deficient affect and serial thinking in conventional AI. Rather than claiming full AGI, it examines a structural approach toward AGI through affective-temporal cognitive integration. Existing AI shows two core limitations: ethical misalignment caused by deficient emotional resonance and catastrophic forgetting caused by limited temporal awareness. To address these issues, JRESA employs TRP as its core computing framework. TRP coordinates logic and emotion through Meta Consciousness Synchronization, while TRP-based ECR supports time-memory interference control and ontological drift management. Simulation results showed identity continuity, with 99.8% context preservation accuracy and 0.8 s autonomous loop restart performance. By shifting from Deep Learning to Flow Learning and establishing TPL, this study suggests a value-aligned and temporally continuous cognitive architecture for future AGI research.

키워드