We prove a joint stochastic zeroth-order lower bound for Goldstein stationarity on a Euclidean query ball, even when the ball is guaranteed to contain a stationary point. In dimension $d$, let $f=\mathbb{E}[F(\cdot;\xi)]$, assume $\mathbb{E}[\operatorname{Lip}(F(\cdot;\xi))^2]\le L_0^2$, and bound the initial objective...
We establish tight randomized higher-order oracle complexity for finding first-order stationary points of nonconvex finite sums. Let $n$ be the number of components, $\Delta>0$ the initial objective-gap bound, $L_p>0$ an individual $p$-th derivative Lipschitz bound, and $\epsilon>0$ the target gradient norm. For every...
Wen-Dao Wu, Hai-Han Zhang, Chen-Heng Zhang et al.· 2 citations
We characterize the fresh-gradient oracle complexity of smooth nonconvex-strongly-concave minimax optimization, with matching upper and lower bounds up to logarithmic factors. Let $\Phi(x)=\max_y f(x,y)$, where $f$ is jointly $L$-smooth and $\mu$-strongly concave in $y$ on unconstrained Euclidean domains, and set $\kap...
Wen-Dao Wu, Hai-Han Zhang, Chen-Heng Zhang et al.· 2 citations
The proof extends adaptive Wishart deferred decisions to a rectangular factor with a $k$-dimensional nullspace while persistence of small compression eigenvalues controls growing $p$ and the spectral endpoint.
The same accuracy exponent holds beyond tensor update rules, even for randomized queries and arbitrary feasible outputs, for smooth convex--concave minimax optimization.
Yan-Yi Li, Hai-Han Zhang, Chen-Heng Zhang et al.· 0 citations
This work reframe knowledge editing from a manifold perspective, viewing it as a localized displacement of an edit sub-manifold within the global knowledge manifold, and proposes ManiEdit, a novel manifold-aware autoregressive editing framework consisting of two core components: Pivot Localization and Manifold-Aware Pr...
Rui Liu, Chen-Heng Zhang, Hao-Xuan Li et al.· 0 citations
A matching $\Omega_p(Q^{2/(3p-1)})$ lower bound is proved for arbitrary adaptive deterministic algorithms and randomized algorithms with per-instance success probability at least $2/3$, without span or tensor-update restrictions.
A single fixed smooth convex hard instance is constructed using a Moreau-smoothed biased max chain, an exact prefix-shielding mechanism, and batched delayed rotations to preserve consistency with the full adaptive transcript and establish the optimality of the square-root complexity branch for deterministic bounded-que...
Wen-Dao Wu, Hai-Han Zhang, Chen-Heng Zhang et al.· 0 citations
This work proves the first near-optimal lower bound for arbitrary adaptive randomized algorithms throughout both accuracy regimes of exact value Lipschitz convex optimization, and develops a posterior mean energy method for adaptive exact max observations.
Haihan Zhang, Chen-Heng Zhang, Zhiquan Qi et al.· 1 citation
Inspired by the human brain, which balances plasticity and stability through complementary episodic storage and gradual consolidation, UniMem is proposed, a self-routing framework for autonomous memory management that consistently outperforms baselines while maintaining execution fidelity.