AI coherence · self-repair · long-horizon agents

Research into how intelligent systems stay coherent.

My research asks how AI systems detect drift, repair internal contradictions, verify recovery and remain stable across long tasks.

Open CRepair explainerORCID

Start here

CRepair

Interactive microsite for detect → repair → verify → stability.

Cross-model replication

Structural self-repair across three LLM families.

Coherent artificial cognition

Structural-dynamical framework and A-Test protocol.