cs.LGSep 3, 2026

Restricted Eigenvalues Beyond Gaussian Width: Threshold Occupancy under Heavy Tails

Authors: Shi FuHuibo XuQixin ZhangDacheng Tao

Abstract

Restricted eigenvalue (RE) bounds govern stable recovery by norm-regularized estimators. For isotropic sub-Gaussian measurements, the benchmark sample size is 1+w(A)21+w(A)^2, where w(A)w(A) is the Gaussian width of the normalized descent cone. The COLT 2015 open-problem note (Banerjee et al., 2015) asked whether the same law follows for heavy-tailed designs from a uniform small-ball condition alone. We give an explicit and systematic negative answer to the general question as formulated there: the proposed law fails in its full dimension-free, arbitrary-set form, and the missing obstruction is simultaneous threshold occupancy. A constant-width polyhedral descent cone with fixed small-ball constants has zero empirical RE on every sample path up to half the ambient dimension. More generally, every finite range space admits exact threshold encoding in an arbitrarily narrow spherical cap and a lift to a full polyhedral descent-cone section. For every fixed threshold VC dimension dd, as β0β\downarrow0, the sharp worst-case sample complexity is Θ(β1[dlog(1/β)+log(1/δ)])Θ(β^{-1}[d\log(1/β)+\log(1/δ)]). The separation persists under exact isotropy and all finite moments: on the same constant-width cone, Gaussian measurements succeed with O(1+log(1/δ))O(1+\log(1/δ)) samples, whereas an isotropic heavy-tailed design fails pathwise for np/logpn\lesssim\sqrt{p/\log p}. Gaussian smoothing yields an everywhere-positive CC^\infty density while retaining arbitrarily poor RE. Under isotropy, a distribution-free fallback governed by affine dimension times squared enclosing radius is sharp on this family.

Explore similar work

Sep 1, 2025stat.ML

The Price of Sparsity: Sufficient Conditions for Sparse Recovery using Sparse and Sparsified Measurements

We consider the problem of support recovery for sparse binary signals from noisy linear measurements. For sparse Gaussian measurement matrices we identify sufficient conditions on the minimal sample size for maximum-likelihood recovery in the high-SNR regime ds/pds/p \to \infty, where pp denotes the signal dimension, ss the number of non-zero components of the signal, and dd the expected number of non-zero components per row of measurement. Combined with known lower bounds, this yields an information-theoretic threshold of order slog(p/s)/log(ds/p)s\log(p/s) / \log(ds/p), making explicit the price of measurement sparsity. In particular, we highlight a regime where the sample-complexity loss from measurement sparsity is logarithmic while the computational gain is nearly linear. Second, we study recovery after sparsifying an originally dense Gaussian design: the observations are generated from the dense design, while estimation uses an independently sparsified design and a rescaled response. In the proportional regime s=αps=αp, d=ψpd=ψp, we prove that, for every fixed target error level δδ and every slack ε>0\varepsilon>0, a sample size of order p/ψ2p/ψ^2 is sufficient for support recovery for arbitrarily small ψψ.
Youssef Chaabouni, David Gamarnik
Jun 12, 2026math.OC

Free Heavy-Tailed Lunch for Muon: A Theoretical Justification of Empirical Success

Non-Euclidean optimisation methods with matrix-valued updates, such as Muon and Scion, have recently shown strong empirical performance for training Transformer models, yet their theoretical advantages over Euclidean methods remain poorly understood. We address this gap in the heavy-tailed non-convex regime, where stochastic gradients have bounded pp-th central moments, p(1,2]p \in (1,2]. We show that certain non-Euclidean methods achieve optimal sample complexity under stronger stationarity measures, while Euclidean methods incur additional dimension-dependent costs. As a consequence, for m×nm \times n matrices, Muon finds an ε\varepsilon-stationary point in nuclear norm within O(min{m,n}Δ1Lε2(σε)pp1)\mathcal{O}\left(\min\{m, n\} \frac{Δ_1 L}{\varepsilon^2} \left(\frac σ\varepsilon \right)^{\frac p {p-1}}\right) samples, absorbing heavy-tailed noise without extra dimension dependence, unlike Euclidean methods. We further prove this sample complexity, including its dimension dependence, is optimal for all first-order methods under nuclear-norm stationarity. Experiments on large language models support our theory. Surprisingly, our results suggest that other Schatten geometries beyond the spectral geometry of Muon can perform competitively in certain settings.
Florian Hübler, Thomas Pethick, Suvrit Sra
May 22, 2026cs.LG

Optimal Dimension-Free Sampling for Regularized Classification

We prove optimal sampling bounds achieving (1±ε)(1\pm\varepsilon)-relative error for a broad class of Lipschitz continuous classification loss functions under various regularization terms. This includes important functions such as logistic and sigmoid loss, hinge loss, and ReLU loss, as prominent and popular representative examples. In particular, we prove k2/ε2k^2/\varepsilon^2 upper and lower bounds for 2/k\|\cdot\|_2/k regularization, and k/ε2k/\varepsilon^2 upper and lower bounds for 1/k\|\cdot\|_1/k regularization. For 22/k\|\cdot\|_2^2/k regularization, the sampling complexity depends mainly on a bounded derivative property: if g(x)g(x)|g'(x)|\leq g(x), and g(0)>0g(0)>0, and gg is monotonic or convex, then it admits linear in kk sampling complexity; otherwise the general bound is k2/ε2k^2/\varepsilon^2. However, if g(0)=0g(0)=0, our results indicate that no dimension-free bounds are possible, and even sublinear bounds are ruled out. All upper bounds are complemented by matching lower bounds up to polylogarithmic terms. Moreover, our work relies conceptually and algorithmically on simple uniform or (squared) norm sampling and hereby improves over recent cubic k3/ε2k^3/\varepsilon^2 sensitivity sampling bounds of (Alishahi and Phillips, ICML'24). This is achieved by refined arguments involving higher moment bounds and empirical process analyses to avoid overcounting that appears in the de-facto standard VC-dimension and sensitivity framework.
Meysam Alishahi, Alexander Munteanu, Simon Omlor +1