Noriyuki Suzuki

Beyond Single-Proof Assurance: Multi-Angle Closure Qualification for AI Safety

AI-safety claims are increasingly supported by heterogeneous evidence, including formal verification, behavioral evaluation, adversarial testing, benchmark scores, independent implementations, and runtime monitoring. Yet accumulating successful evaluations does not by itself esta...

Beyond Scaling IV: World Models Without Future Preservation / Meta-Complex Observation and the Changing Structure of Reachable Futures

This paper examines a limitation of predictive and agentic world models: the distinction between predicting future states and observing how an agent's own intervention changes the structure of reachable futures. World models can predict action-conditioned trajectories, and existi...