{"id":"arxiv/2603.19465","url":"https://arxiv.org/abs/2603.19465","tags":["cs.LG","cs.AI","math.OC"],"score":0,"title":"Global Convergence of Multiplicative Updates for the Matrix Mechanism: A Collaborative Proof with Gemini 3","domain":"cs.LG","source":"arxiv","authors":["Keith Rush"],"reviews":0,"subject":"cs.LG","summary":"We analyze a fixed-point iteration $v \\leftarrow φ(v)$ arising in the optimization of a regularized nuclear norm objective involving the Hadamard product structure, posed in DMR+22 in the context of an optimization problem over the space of algorithms in private machine learning. We prove that the iteration $v^{(k+1)} = \\text{diag}((D_{v^{(k)}}^{1/2} M D_{v^{(k)}}^{1/2})^{1/2})$ converges monotonically to the unique global optimizer of the potential function $J(v) = 2 \\text{Tr}((D_v^{1/2} M D_v^{1/2})^{1/2}) - \\sum v_i$, closing a problem left open there. The bulk of this proof was provided…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-03-19","reactions":{}}