3 papers
cs.GT2026
A Pigouvian Matchmaker Mechanism for De-escalating the AGI Race
Eduard Kapelko
A formal mechanism is presented in which a willing regulator-matchmaker fosters cooperation on resources among participants in the AGI race, collects a Pigouvian tax based on the s…
cs.LG2026
Active Inference Agency Formalization, Metrics, and Convergence Assessments
Eduard Kapelko
This paper addresses the critical challenge of mesa-optimization in AI safety by providing a formal definition of agency and a framework for its analysis. Agency is conceptualized…
cs.CL2025
Cyclic Ablation: Testing Concept Localization against Functional Regeneration in AI
Eduard Kapelko
Safety and controllability are critical for large language models. A central question is whether undesirable behaviors like deception are localized functions that can be removed, o…