go-$m$HC: Direct Parameterization of Manifold-Constrained Hyper-Connections via Generalized Orthostochastic Matrices

• URL: http://arxiv.org/abs/2604.02309v1
• Date: Thu, 02 Apr 2026 17:51:19 GMT
• Title: go-$m$HC: Direct Parameterization of Manifold-Constrained Hyper-Connections via Generalized Orthostochastic Matrices
• Abstract summary: We introduce a novel exact parameterization grounded in the theory of generalized orthostochastic matrices.<n>We instantiate this parameterization in go-$m$HC.<n>Our method composes naturally with Kronecker-factorized methods, substantially recovering expressivity at similar FLOP costs.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Doubly stochastic matrices enable learned mixing across residual streams, but parameterizing the set of doubly stochastic matrices (the Birkhoff polytope) exactly and efficiently remains an open challenge. Existing exact methods scale factorially with the number of streams ($d$), while Kronecker-factorized approaches are efficient but expressivity-limited. We introduce a novel exact parameterization grounded in the theory of generalized orthostochastic matrices, which scales as $\mathcal{O}(d^3)$ and exposes a single hyperparameter $s$ which continuously interpolates between a computationally efficient boundary and the fully expressive Birkhoff polytope. Building on Manifold-Constrained Hyper-Connections ($m$HC), a framework for learned dynamic layer connectivity, we instantiate this parameterization in go-$m$HC. Our method composes naturally with Kronecker-factorized methods, substantially recovering expressivity at similar FLOP costs. Spectral analysis indicates that go-$m$HC fills the Birkhoff polytope far more completely than Kronecker-factorized baselines. On synthetic stream-mixing tasks, go-$m$HC achieves the minimum theoretical loss while converging up to $10\times$ faster. We validate our approach in a 30M parameter GPT-style language model. The expressivity, efficiency, and exactness of go-$m$HC offer a practical avenue for scaling $d$ as a new dimension of model capacity.

Related papers

• Low-Rank Dependence Decomposition via Accelerated Symmetric Non-negative Matrix Factorization [0.0]
  We present a large-scale GPU study of seven algorithm families (over 30 configurations) on absolute Pearson correlation and tail pairwise dependence matrices.<n>A trace-identity reformulation eliminates all $n times n$ intermediates, so a single GPU reaches $n approx 105$ and multi-node distribution scales to $n = 106$ and beyond.
  arXiv  Detail & Related papers  (2026-07-27T14:56:00Z)
• Sharper Analysis of Single-Loop Methods for Bilevel Optimization [81.36947969465848]
  We establish sharper convergence results for single-loop approximate implicit differentiation (AID) and iterative differentiation (ITD) methods.<n>We prove that the error is $mathcalO(2)$, exactly matching the known lower bound and improving upon the previous $mathcalO(3)$ guarantee.
  arXiv  Detail & Related papers  (2026-07-11T11:41:42Z)
• TBP-mHC: full expressivity for manifold-constrained hyper connections through transportation polytopes [0.0]
  Hyper-Connections (HC) improve residual networks by introducing learnable mixing across multiple residual streams.<n>We propose Transportation Birkhoff Polytope (TBP) parameterizations and their Recursive variants (RTBP)<n>Our approach avoids iterative normalization and explosion while preserving full expressivity of the Birkhoff polytope.
  arXiv  Detail & Related papers  (2026-05-20T20:31:10Z)
• Regularized Online RLHF with Generalized Bilinear Preferences [68.44113000390544]
  We consider the problem of contextual online RLHF with general preferences.<n>We adopt the Generalized Bilinear Preference Model to capture preferences via low-rank, skew-symmetric matrices.<n>We prove that the dual gap of the greedy policy is bounded by the square of the estimation error.
  arXiv  Detail & Related papers  (2026-02-26T15:27:53Z)
• Parameter-free Algorithms for the Stochastically Extended Adversarial Model [59.81852138768642]
  Existing approaches for the Extended Adversarial (SEA) model require prior knowledge of problem-specific parameters, such as the diameter of the domain $D$ and the Lipschitz constant of the loss functions $G$.<n>We develop parameter-free methods by leveraging the Optimistic Online Newton Step (OONS) algorithm to eliminate the need for these parameters.
  arXiv  Detail & Related papers  (2025-10-06T10:53:37Z)
• Inertial Quadratic Majorization Minimization with Application to Kernel Regularized Learning [1.0282274843007797]
  We introduce the Quadratic Majorization Minimization with Extrapolation (QMME) framework and establish its sequential convergence properties.<n>To demonstrate practical advantages, we apply QMME to large-scale kernel regularized learning problems.
  arXiv  Detail & Related papers  (2025-07-06T05:17:28Z)
• ALoRE: Efficient Visual Adaptation via Aggregating Low Rank Experts [71.91042186338163]
  ALoRE is a novel PETL method that reuses the hypercomplex parameterized space constructed by Kronecker product to Aggregate Low Rank Experts.<n>Thanks to the artful design, ALoRE maintains negligible extra parameters and can be effortlessly merged into the frozen backbone.
  arXiv  Detail & Related papers  (2024-12-11T12:31:30Z)
• Orthogonal Directions Constrained Gradient Method: from non-linear equality constraints to Stiefel manifold [16.099883128428054]
  We propose a novel algorithm, the Orthogonal Directions Constrained Method (ODCGM) ODCGM only requires computing a projection onto a vector space. We show that ODCGM exhibits the near-optimal oracle complexities.
  arXiv  Detail & Related papers  (2023-03-16T12:25:53Z)
• CEDAS: A Compressed Decentralized Stochastic Gradient Method with Improved Convergence [9.11726703830074]
  In this paper, we consider solving the distributed optimization problem under the communication restricted setting. We show the method over com-pressed exact diffusion termed CEDAS" In particular, when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when when
  arXiv  Detail & Related papers  (2023-01-14T09:49:15Z)
• Permutation Compressors for Provably Faster Distributed Nonconvex Optimization [68.8204255655161]
  We show that the MARINA method of Gorbunov et al (2021) can be considered as a state-of-the-art method in terms of theoretical communication complexity. Theory of MARINA to support the theory of potentially em correlated compressors, extends to the method beyond the classical independent compressors setting.
  arXiv  Detail & Related papers  (2021-10-07T09:38:15Z)
• Reducing the Variance of Gaussian Process Hyperparameter Optimization with Preconditioning [54.01682318834995]
  Preconditioning is a highly effective step for any iterative method involving matrix-vector multiplication. We prove that preconditioning has an additional benefit that has been previously unexplored. It simultaneously can reduce variance at essentially negligible cost.
  arXiv  Detail & Related papers  (2021-07-01T06:43:11Z)
• An Online Riemannian PCA for Stochastic Canonical Correlation Analysis [37.8212762083567]
  We present an efficient algorithm (RSG+) for canonical correlation analysis (CCA) using a reparametrization of the projection matrices. While the paper primarily focuses on the formulation and technical analysis of its properties, our experiments show that the empirical behavior on common datasets is quite promising.
  arXiv  Detail & Related papers  (2021-06-08T23:38:29Z)
• Unified Convergence Analysis for Adaptive Optimization with Moving Average Estimator [75.05106948314956]
  We show that an increasing large momentum parameter for the first-order moment is sufficient for adaptive scaling.<n>We also give insights for increasing the momentum in a stagewise manner in accordance with stagewise decreasing step size.
  arXiv  Detail & Related papers  (2021-04-30T08:50:24Z)
• A conditional one-output likelihood formulation for multitask Gaussian processes [0.0]
  Multitask Gaussian processes (MTGP) are the Gaussian process framework's solution for multioutput regression problems. Here we introduce a novel approach that simplifies the multitask learning. We show that it is computationally competitive with state of the art options.
  arXiv  Detail & Related papers  (2020-06-05T14:59:06Z)
• Multi-Objective Matrix Normalization for Fine-grained Visual Recognition [153.49014114484424]
  Bilinear pooling achieves great success in fine-grained visual recognition (FGVC) Recent methods have shown that the matrix power normalization can stabilize the second-order information in bilinear features. We propose an efficient Multi-Objective Matrix Normalization (MOMN) method that can simultaneously normalize a bilinear representation.
  arXiv  Detail & Related papers  (2020-03-30T08:40:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.