Answers are grounded in this paper and cite their sources.
We compute exact projection comparison and finite-decomposition bounds in Rørdam's simple separable nuclear \(C^*\)-algebra \(G\), which contains both finite and infinite projections. Write \(H_S\) for the tensor product of the coordinate Hopf lines indexed by a finite set \(S\), viewed as a coefficient-stage projection in \(G\). For nonempty finite \(S,J\) in a common coordinate layer \(Λ_r\), \(r≥1\), we prove \[ min\{k≥1:H_J H_S⊕ k\}=1+min\{|S|,|S J|\}. \] Here \(S J\) is the symmetric difference and \(\) denotes Murray–von Neumann subequivalence. The same number is the least \(k\) for which \([H_J]\) is a finite sum of projection classes represented in \(M_k(H_SGH_S)\), with equivalences taken in the ambient stabilization. We exhibit a projection class below \(2[H_S]\) that is no finite sum of classes bounded by \([H_S]\). Consequently, the stabilized projection monoid \(V(G)\) fails Riesz decomposition and refinement; the witnessing projection \(2H_S\) is finite when \(|S|≥2\). The monoids generated at successive matrix sizes increase strictly through size \(|S|+1\); the same bounds give exact cutoffs for representation by unital approximately finite-dimensional subalgebras. The formula and the nonrefinement obstruction hold in the original multiplier-algebra inductive limit without the layer restriction. GitHub repository: eulogica/projection-decomposition-rordam
No takes yet. Share an insight, caveat, or question.
Martino Lupini (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: