MCCB Experiment Code and Reproducible Results

This archive contains the experimental code and generated resultsfor the study: "Multi-Channel Conscious Binding (MCCB): Dynamic Concept Steeringin Frozen Large Language Models". The repository provides a minimal, all-in-one implementation of theMCCB method, including concept vector construction, calibration,activation measurement, and visualization. Contents:- A single Python script reproducing all experiments- Generated numerical results (CSV)- Visualization figures (PNG)- A human-readable summary (TXT) The code is designed for clarity and reproducibility rather thanperformance optimization. This archive is intended to support reproducibility, inspection,and further extension of the MCCB framework.