Skip to main content

Research

Exploring a Foundational Computational Framework for Cognitive Enhancement and Artificial General Intelligence

KeplerJAI Research is dedicated to computational modelling that reveals the boundaries and dynamical mechanisms of human cognition under accelerated information input, and to providing a testable assessment framework for the grading and evaluation of AGI.

A Computational Framework for Cognitive Enhancement

This study proposes a falsifiable computational framework in which cognitive load is modelled as a bounded accumulation process. At low input rates, the error rate grows approximately linearly; beyond the overload threshold, working-memory overflow triggers a power-law transition. The framework provides testable boundaries for safe cognitive enhancement and for BCI-based cognitive regulation.

Read the Full Paper (Zenodo)

Human–AI Alignment and BCI Regulation

Building on the dynamical model above, the research further proposes two downstream application directions: Human–AI Structured Knowledge Transfer (based on functional alignment, rejecting the geometric isomorphism assumption) and BCI-based closed-loop cognitive-load regulation. These applications aim to adjust the rate of information input dynamically through real-time neural monitoring, thereby achieving safe cognitive enhancement.

Read the Full Paper (Zenodo)

Published Papers

Download All Papers on Zenodo