{"id":"arxiv/2610.16797","url":"https://arxiv.org/abs/2610.16797","tags":["cs.IT","cond-mat.dis-nn","math.ST"],"score":0,"title":"Information-Theoretic Limits for Rectangular Spiked Matrix Models with Rotationally Invariant Noise","domain":"cs.IT","source":"arxiv","authors":["Qun Chen","Rudy Skerk","Jean Barbier","Francesco Camilli","Junjie Ma","Haochuan Zhang"],"reviews":0,"subject":"cs.IT","summary":"We study the information-theoretic limits of inference in rectangular rank-one spiked matrix models with additive rotationally invariant noise, thereby going beyond the standard assumption of independent noise entries. Using the replica method from statistical physics, we derive a variational formula for the mutual information between the signal and the data, together with a system of saddle point equations predicting the minimum mean squared error in estimating the rank-one spike. We further show that, in the Bayes-optimal setting, these saddle point equations coincide with the fixed point…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-10-12","reactions":{}}