A linearization map for genuine equivariant algebraic $K$-theory

Basic Information

• Paper ID: 2309.08025
• Title: A linearization map for genuine equivariant algebraic $K$-theory
• Authors: Maxine E. Calle, David Chan, Andres Mejia
• Classification: math.AT (Algebraic Topology), math.KT (K-Theory and Homology)
• Publication Date: November 19, 2025 (v3)
• Paper Link: https://arxiv.org/abs/2309.08025

Abstract

This paper introduces an equivariant version of algebraic $K$-theory for coefficient systems of rings, taking values in genuine $G$-spectra for a finite group $G$. The authors use this construction to establish genuine $G$-spectra $K_G(\mathbb{Z}[\pi_1(X)])$ associated with $G$-spaces $X$, providing a natural home for equivariant versions of classical invariants such as Wall finiteness obstruction and Whitehead torsion. The paper establishes a comparison between this $K$-theory spectrum and the equivariant $A$-theory of Malkiewich-Merling through a genuine equivariant linearization map.

Research Background and Motivation

1. Research Problem

Classical algebraic $K$-theory $K(\mathbb{Z}[\pi_1(X)])$ encodes important geometric invariants of topological spaces $X$, including Euler characteristic, Wall finiteness obstruction, and Whitehead torsion. When the space $X$ admits an action by a finite group $G$, the group ring $\mathbb{Z}[\pi_1(X)]$ inherits a $G$-action, and its $K$-theory spectrum becomes a naïve $G$-spectrum. However, this naïve structure is insufficient to fully capture the $G$-homotopy type of $X$.

2. Importance of the Problem

• Missing equivariant invariants: Equivariant analogues of Wall finiteness obstruction and Whitehead torsion are known to exist (Baglivo 1979, Andrzejewski 1986, Illman 1974, etc.), but lack a unified framework at the spectral level
• Necessity of genuine spectra: To fully describe equivariant homotopy types, genuine $G$-spectra rather than naïve $G$-spectra are required
• Connection to equivariant $A$-theory: The genuine equivariant $A$-theory $A_G(X)$ constructed by Malkiewich-Merling needs to be connected to algebraic $K$-theory

3. Limitations of Existing Approaches

• Merling's construction: For rings $R$ with $G$-action, Merling constructed genuine $G$-spectra $K^\theta(R)$, but when $|G|$ is invertible in $R$, the fixed points $K^\theta(R)^H \simeq K(R^H_\theta[W_GH])$ only capture the fixed points of the ring, while $(Z[\pi_1(X)])^H \neq Z[\pi_1(X^H)]$, failing to accurately describe the data of space fixed points
• Lück's approach: The low-dimensional homotopy groups of Lück's constructed spectrum contain equivariant invariants, but the relationship to genuine $G$-spectra remains unclear

4. Research Motivation

This paper adopts the Elmendorf philosophy: equivariant homotopy theory can be understood by studying the collection of fixed point spaces $\{X^H\}_{H\leq G}$ and their restriction maps. This motivates the idea of constructing genuine $G$-spectra from coefficient systems of rings (presheaves on the orbit category).

Core Contributions

1. $K$-theory of coefficient rings: Constructs a functor $K_G(S)$ from coefficient systems of rings to genuine $G$-spectra, whose fixed points have an explicit splitting (Theorem A): $K_G(S)^G \simeq \prod_{(H)\leq G} K(S^H_\theta)$
2. $K$-theory of equivariant fundamental groups: For $G$-spaces $X$, defines coefficient rings $\mathbb{Z}[\pi_1(X)]$, obtaining genuine $G$-spectra $K_G(\mathbb{Z}[\pi_1(X)])$, recovering Lück's results and providing spectral-level realizations of equivariant invariants
3. Genuine equivariant linearization map: Constructs a map $L: A_G(X) \to K_G(\mathbb{Z}[\pi_1(X)])$ (Theorem C), proving it is equivariantly 2-connected and compatible with fixed point splittings (Theorem D)
4. Realization of geometric invariants: Proves that equivariant Euler characteristic, Wall finiteness obstruction, and Whitehead torsion are realized as elements in homotopy groups of $A_G(X)$ (Theorem B)
5. Technical contributions: Develops a systematic method for constructing spectral Mackey functors and their morphisms from 2-category inputs (Appendix B), expected to have independent applications

Detailed Methodology

Task Definition

Core task: For a finite group $G$ and a coefficient system of rings $S: \mathcal{O}_G^{op} \to \text{Ring}$, construct a genuine $G$-spectrum $K_G(S)$ such that:

• The value at each subgroup $H\leq G$ is $K(P^H_S)$ ($K$-theory of the category of finitely generated projective modules)
• Transfer, restriction, and conjugation maps are induced by induction, restriction functors and their $K$-theory inductions
• Fixed points have a computable splitting

Input:

• Coefficient rings $S: \mathcal{O}_G^{op} \to \text{Ring}$, i.e., for each $G/H$ a ring $S^H$ equipped with restriction maps and Weyl group actions
• Or more specifically, coefficient rings $\mathbb{Z}[\pi_1(X)](G/H) := \mathbb{Z}[\pi_1(X^H)]$ defined by $G$-spaces $X$

Output: Genuine $G$-spectrum (via spectral Mackey functor model)

Model Architecture

1. Categorical Structure of Coefficient Systems

Definition (2.1-2.2):

• A coefficient system $M: \mathcal{O}_G^{op} \to \text{Ab}$ is a functor from the orbit category to abelian groups
• A coefficient ring $S$ is a coefficient system with values in the category of rings, with restriction maps and Weyl group actions being ring homomorphisms

Key structure:

• Monoidal product $\square$: $(M\square N)^H = M^H \otimes N^H$, with unit the constant $\mathbb{Z}$ coefficient system
• Coefficient rings are precisely monoids in this monoidal category
• The category of $S$-modules $\text{Mod}_S$ is defined as modules in this monoidal category

Twisted group ring (2.6-2.7): For a ring $R$ with $G$-action, the twisted group ring $R_\theta[G]$ has multiplication: $(r_1g_1)(r_2g_2) = (r_1r_2^{g_1})g_1g_2$ An $S$-module $M$ is equivalent to having $M^H$ be an $S^H_\theta[W_GH]$-module at each level, compatible with restriction maps (Lemma 2.9)

2. Induction and Restriction Functors

For $H\leq K$, define a functor $\rho: \mathcal{O}_H \to \mathcal{O}_K$:

• Restriction functor $R^K_H: \text{Coeff}_K \to \text{Coeff}_H$ is $(ρ^{op})^*$
• Induction functor $I^K_H: \text{Coeff}_H \to \text{Coeff}_K$ is the left Kan extension $ρ^{op}_!$

Computation formula (Lemma 2.18): $I^K_H(M)^J \cong \bigoplus_{x\in (K/H)^J} M^{J^x}$ where $J^x = x^{-1}Jx$ is the conjugate subgroup

Frobenius reciprocity (Lemma 2.16): $M \square I^K_H(N) \cong I^K_H(R^K_H(M) \square N)$

Mackey double coset formula (Proposition 2.26): $R^K_H I^K_J \cong \bigoplus_{\gamma \in J\backslash K/H} I^H_{H\cap J^\gamma} R^{J^\gamma}_{H\cap J^\gamma} c_\gamma$

These functors restrict to the category of projective modules, giving adjunctions: $\text{Proj}_{R^G_K(S)} \rightleftarrows \text{Proj}_{R^G_H(S)}$

3. From Categorical Mackey Functors to Spectral Mackey Functors

2-category framework (Section 3 and Appendix A):

• Define a strict 2-category $\mathbf{B}_G$ with objects being subgroups of $G$, and morphism categories $\mathbf{B}_G(H,K)$ essentially consisting of formal transfers, restrictions, and conjugations
• Key technical result (Proposition 3.3, generalizing Malkiewich-Merling): Given a strict 2-functor $F: \mathbf{B}_G \to \text{Wald}$ satisfying condition $(\star)$ (preserving coproducts and zero objects), one can construct a spectral Mackey functor $K_G(F)$

Condition $(\star)$: For any $A\in F(H)$ and $S,T\in \mathbf{B}_G(H,K)$: $F(S\sqcup T)(A) \cong F(S)(A) \vee F(T)(A), \quad F(\emptyset)(A) \cong 0_K$

Handling pseudofunctors (Proposition 3.4): Actually only pseudofunctors (compositions satisfy up to isomorphism) are needed; via standard strictification Proposition 3.3 applies

4. Construction of $K_G(S)$

Waldhausen category of projective modules: For coefficient rings $S$, $P^H_S$ is the category of finitely generated projective modules over $R^G_H(S)$, equipped with exact structure (admissible monics as monomorphisms, split epimorphisms as epimorphisms)

Pseudofunctor construction (Proposition 3.6): The assignment $H \mapsto P^H_S$ equipped with induction, restriction, and conjugation functors assembles into a pseudofunctor $\mathbf{B}_G \to \text{Wald}$ satisfying $(\star)$

Proof outline:

• Pseudofunctoriality of induction and restriction comes from the Mackey formula (2.26) and conjugation properties (2.25)
• Condition $(\star)$ is guaranteed by the coproduct properties of coefficient systems
• Complete proof deferred to Appendix B (Proposition B.43), using systematic methods

Statement of Theorem A (Theorems 3.1, 3.7): $K_G(S)^G \simeq K(\text{Proj}_S) \simeq \prod_{(H)\leq G} K(S^H_\theta)$ where the product ranges over conjugacy class representatives, $S^H_\theta = S(G/H)_\theta[W_H]$

Technical Innovations

1. Breakthrough in Coefficient System Perspective

• Innovation: Rather than directly handling rings with $G$-action, systematically use coefficient systems (presheaves on the orbit category)
• Advantage: Naturally encodes fixed point information at all subgroups, avoiding the problem $(R^G)^H \neq R^H$ in Merling's method
• Justification: Aligns with Elmendorf philosophy—equivariant homotopy theory is determined by fixed point systems

2. Fine Control of Splitting Theorem

• Filtration method (Section 3.1): Define subcategories $P_{\leq i}$ (vanishing on $G/H_j$, $j>i$) and $P_i$ ($H_i$-generated projective modules)
• Key lemma (3.12): Each $P\in P_{\leq i}$ admits a unique split short exact sequence $P_i \to P \to P/P_i$
• Equivalence (3.15): Construct functor $\Phi_i: \text{Proj}_{S^{H_i}_\theta} \to P_i$, proving the evaluation functor $\text{ev}_{G/H_i}: P_i \to \text{Proj}_{S^{H_i}_\theta}$ is an equivalence
• Result: Obtain split fiber sequences, inductively producing the splitting

3. Perfect Complexes Perspective

• Motivation: Linearization maps are more naturally defined at the chain complex level
• Construction (Section 3.2): Define the category $\text{Perf}_{\leq i}$ of $H_i$-bounded perfect complexes
• Key proposition (3.24): There exists a homotopy equivalence $K(\text{Perf}_{\leq i}) \simeq K(\text{Perf}_{\leq(i-1)}) \times K(\text{Perf}_{S^{H_i}_\theta})$ with projections induced by evaluation $\text{ev}_{G/H_i}$
• Advantage: Allows defining linearization at the complex level while preserving splitting structure

4. Construction Strategy of Linearization Map

Functor definition (Section 4.1): For $Y\in R^G_{fd}(X)$ (equivariant finitely dominated retractive space), define $L_G(Y)(G/H) = C_*(\tilde{Y}^H, \tilde{X}^H)$ where $\tilde{Y}^H$ is the pullback of $Y^H$ along the retraction $Y^H\to X^H$

Weak exactness: $L_G$ does not preserve coproducts (only up to quasi-isomorphism), but preserves weak pushouts, thus inducing $K$-theory maps (using Blumberg-Mandell's $S'_\bullet$-construction)

Handling pseudo-naturality:

• $L$ is not a pseudo-natural transformation (transfers only preserve up to quasi-isomorphism)
• Introduce "weak natural transformation" concept (lax transformation with 2-cells being weak equivalences)
• Appendix B proves this suffices to induce genuine $G$-spectrum maps (Theorem B.12)

Compatibility with splitting (Theorem 4.20): Key commutative diagram