Havenlon Research
Research on execution control, execution boundaries, and adversarial completeness for autonomous systems.
An Authority-Decomposition Framework for High-Risk Automated Systems
An action-relative framework for identifying which trust-domain coalitions can cause protected execution in high-risk automated systems.
AuthorsMengting Wu, Lin Wang, Yong Zhang
Execution Boundary Language Core
An execution-boundary conformance profile for deciding when a fully materialized AI-generated action may receive execution authority.
AuthorsMengting Wu, Lin Wang, Yong Zhang, Jiang Deng
Scope, Obligations, and Cutoff Reasoning
An analytical model for determining whether the declared evidence obligations of an intent-bound execution are completely covered within a defined scope and assessment cutoff.
AuthorsMengting Wu, Lin Wang, Yong Zhang, Jiang Deng
