Relative Invariants from Moving Frames on an Extended Manifold

Leonid Bedratyuk · 2026

Read Relative Invariants from Moving Frames on an Extended Manifold by Leonid Bedratyuk (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-06660.txt.

Introduction

Leonid Bedratyuk builds a new way to make group-based yardsticks in a 2026 paper. His base is the moving frame method, changed in a clear-cut way. It works on an enlarged space, which gives the method room to run.

A yardstick of the kind he builds shifts by a power of a scale factor. That scale factor comes from the group act and is known in closed form. Old frame tools find fixed yardsticks, but some tasks need this looser kind. His changed method gives steps one can carry out, not just a proof that they exist. It links old work by Cartan to present-day work on group acts. Those who study shape and symmetry gain a tool they can apply at once. New group acts can now be worked through the same steps.

Catalog record

CREATORLeonid Bedratyuk
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE57 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-06660

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Relative Invariants from Moving Frames on an Extended Manifold Leonid Bedratyuk Abstract . A constructive modification of the moving frame method is developed in this paper for the construction of relative invariants of regular Lie group actions. Let a relative invariant I I of weight ω \omega transform according to the rule I ⁡ ( g ⋅ 𝒙 ) = μ ​ ( g , 𝒙 ) ω ​ I ​ ( 𝒙 ) , I(g\cdot\boldsymbol{x})=\mu(g,\boldsymbol{x})^{\omega}I(\boldsymbol{x}), where μ : G × ℳ → ℝ × \mu:G\times\mathcal{M}\to\mathbb{R}^{\times} is a scalar multiplier (1-cocycle). It is shown that the cocycle property of μ \mu is equivalent to the well-definedness of the twisted group action on the extended manifold ℳ ^ = ℳ × ℝ × \widehat{\mathcal{M}}=\mathcal{M}\times\mathbb{R}^{\times} , and that relative invariants on ℳ \mathcal{M} are in one-to-one correspondence with absolute invariants of this action on ℳ ^ \widehat{\mathcal{M}} . The main result is that, given a moving frame, the invariantization of the multiplier is a canonical relative invariant of weight − 1 -1 . This enables the constructive realization of any weight and yields an explicit formula for an arbitrary relative invariant in terms of the fundamental absolute invariants and the invariantized multiplier. Examples are provided to demonstrate the application of the proposed approach for the projective group P ​ G ​ L ​ ( 3 , ℝ ) PGL(3,\mathbb{R}) . Keywords: moving frame algorithm; relative invariants; multiplier; 1 1 -cocycle; twisted action; group cohomology; invariantization. 1. Introduction Let G be an r -dimensional Lie group acting smoothly on an n -dimensional manifold \mathcal{M} . This action induces an action of G on the field \mathcal{F}(\mathcal{M}) of rational functions on \mathcal{M} . A function f\in\mathcal{F}(\mathcal{M}) is called a relative invariant with a smooth multiplier \mu:G\times\mathcal{M}\to\mathbb{R}^{\times} if (1) f(g\cdot\boldsymbol{x})=\mu(g,\boldsymbol{x})\,f(\boldsymbol{x}),\qquad\forall\,g\in G,\ \boldsymbol{x}\in\mathcal{M}. More generally, one can introduce a rational weight \omega\in\mathbb{Q} and replace \mu by the weighted multiplier, i.e. f(g\cdot\boldsymbol{x})=\mu(g,\boldsymbol{x})^{\omega}\,f(\boldsymbol{x}). In this terminology, a relative invariant of weight 0 is an absolute invariant. The classification problem for relative invariants of transformation group actions arises throughout modern mathematics and its applications. Relative invariants occur naturally in the setting of affine and projective transformations, where, for instance, area or volume is no longer preserved but is scaled by the determinant of the transformation matrix. A general description of relative invariants makes it possible to treat a broad range of invariant structures as sections of vector bundles preserved by the induced group action. Relative invariants play a fundamental role in differential geometry (invariant vector fields, differential forms, volume elements) [ 3 , 1 , 4 , 2 ] , in the theory of invariant differential operators [ 5 ] , in the study of invariant metrics and connections [ 6 ] , in number theory (automorphic and modular forms, Epstein zeta functions) [ 8 , 7 ] , in classical mechanics (relative integral invariants of Poincaré–Cartan type and symplectic geometry) [ 9 , 10 ] , and in applied problems such as computer vision [ 11 , 12 , 13 ] , the theory of special functions (in particular, Mellin–Barnes type integral representations, matrix-argument functions, and Whittaker functions) [ 14 , 15 ] , and quantum mechanics (gauge invariance and the transformation properties of wave functions viewed as relative invariants) [ 16 ] . Despite their wide applicability, the systematic theory of relative invariants is far less developed than that of absolute invariants. While the classification of absolute invariants for regular Lie group actions is a direct consequence of the Frobenius theorem, an analogous general theory for relative invariants has only been achieved more recently. In particular, [ 1 , 17 ] completely solves the general classification problem for relative invariants of regular multiplier representations of connected Lie group actions, providing an explicit formula for the precise number of functionally independent relative invariants and establishing their structure. The breadth of applicability is illustrated through a detailed treatment of several important geometric constructions arising as relative invariants of (vector and, in particular, tensor) bundle actions, including invariant vector fields and differential forms, purely geometric conditions for the existence of such fields, and an extensive discussion of applications to invariant connections (including inhomogeneous multiplier representations). In practically important problems, however, a central issue remains the constructive extraction of generating relative invariants within a function class suitable for computation; the present work focuses on this algorithmic aspect. Our goal is to develop a general methodology for constructing relative invariants using the moving frame method—an effective constructive technique originating in the geometric ideas of Cartan and developed in modern form by Fels and Olver [ 18 , 19 ] . The moving frame method has become a standard tool for producing absolute and differential invariants of regular Lie group actions. At the same time, its potential for a systematic treatment of relative invariants has not been fully exploited in the literature. The proposed approach rests on two key ideas that reduce the problem to an explicit constructive algorithm. First, we pass to the extended space \widehat{\mathcal{M}}=\mathcal{M}\times\mathbb{R}^{\times} and introduce a twisted action of G in which the multiplier \mu appears as the factor governing the transformation of the additional coordinate. This reduces the construction of relative invariants on \mathcal{M} to the construction of absolute invariants for the extended action on \widehat{\mathcal{M}} . Second, assuming that a moving frame for the action on M has already been constructed, we show that invariantization of the multiplier yields a canonical relative invariant of weight -1 . In this way, relative invariants emerge as a natural extension of absolute invariants within the unified geometric paradigm of the moving frame method. Our results also generalize and systematize observations that have appeared in concrete examples, in particular for projective groups, where invariantization of the multiplier agrees with natural frame-dependent quantities arising in projectively invariant constructions [ 20 , 21 ] . Nevertheless, the main emphasis of this paper is not on classification or global issues, but on the constructive procedure: given a moving frame, how to efficiently produce relative invariants. We also obtain a complete description of admissible multipliers via a cohomological interpretation. A multiplier is naturally viewed as a 1 -cocycle in the multiplicative bar complex of the group action with coefficients in \mathcal{F}(M)^{\times} , and the cocycle condition is equivalent to the functional cocycle identity for multipliers. Using the triviality of the first cohomology group in this setting, we show that every 1 -cocycle is a 1 -coboundary. Thus, the cohomological approach not only clarifies the geometric meaning of the multiplier as a ‘‘change of gauge’’, but also yields a closed-form classification of all multipliers. In a broader geometric context, one encounters the nontrivial realizability problem: whether a prescribed 1 -cocycle can occur as the multiplier of some relative invariant, notably in connection with invariant divisors and equivariant line bundles; see [ 22 ] . For regular smooth actions in the class C^{\infty}(M) , local results were obtained by Fels and Olver [ 1 ] , whereas in the general case no universal criterion is known. In the present work we remain in a constructive regular setting and in the class of rational functions, where the existence of a moving frame implies the triviality of the first

Read the complete text