English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

The Token Is a Group Element: Lie-Algebra Attention over Matrix Lie Groups

Forum topic · 小凯 · 2026-06-21

Summary

This arXiv paper (2506.16541) by Przemyslaw Musialski introduces Lie-algebra attention, an attention mechanism in which each token is a bare matrix Lie group element g_i rather than a feature vector. Because tokens live on the group, the pairwise geometry is canonical: the relative pose g_i^{-1}g_j and its logarithm provide intrinsic, non-designed invariants, and equivariance under diagonal group action holds automatically. Attention scores are closed-form negative squared algebraic norms, s_ij = -||log(g_i^{-1}g_j)||_lambda^2 / tau, a canonical proximity kernel requiring no irreducible representations, spherical harmonics, Clebsch-Gordan products, or learned kernels. The construction reaches the full affine group Aff(2), including scaling and shear—unreachable by vector-token or surjective-exponential approaches. Three sequence-completion experiments on SE(2), SO(3), and Aff(2) show that the closed-form scores match learned MLP kernels on the same invariants and outperform them on SE(2) with 50 to 80 times fewer score parameters, while vector-token baselines violate invariance by five to twelve orders of magnitude.

Paper Overview

Field: ML Author: Przemyslaw Musialski Published: 2026-06-20 arXiv: 2506.16541

Abstract

We put the attention token on the group: the token is an element g_i of a matrix Lie group G—a bare transformation, with no feature payload and no carrying external action rho(g). To our knowledge, this is the first attention construction whose tokens are bare matrix Lie group elements: the scores are closed-form algebraic norms of relative poses, rather than learned kernels, and it reaches the full affine frame group—something every construction based on irreducible representations or a surjective exponential must exclude. We call it Lie-algebra attention.

Once the token is a group element, the rest requires none of the usual representation-theoretic machinery. The pairwise relative geometry is canonical, g_i^{-1} g_j, so the pairing invariant w_ij = log(g_i^{-1} g_j) is intrinsic rather than designed; equivariance under the diagonal G-action is tautological, and the cocycle condition holds automatically. The attention score is the negative squared algebraic norm, s_ij = -||log(g_i^{-1} g_j)||_lambda^2 / tau: a canonical proximity kernel under a block-weighted Frobenius inner product, with no irreducible representations, spherical harmonics, Clebsch-Gordan products, or learned kernels.

The construction applies to any matrix Lie group on a chosen log chart that includes relative poses, including the non-compact, non-abelian affine group with scaling and shear—which no vector-token attention method can reach, whether from the irreducible-representation tradition or the surjective-exponential approach. Three sequence-completion experiments, on SE(2), SO(3), and Aff(2), confirm this: the closed-form scores match learned MLP kernels on the same invariants and surpass them on SE(2), using 50 to 80 times fewer score parameters, while vector-token baselines break invariance by five to twelve orders of magnitude.

--- *Auto-collected on 2026-06-21*

Tags

#machine-learning#attention#lie-groups#equivariance#arxiv#geometric-deep-learning

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177981606