Convergence and Knowledge Processing in Multi-Agent