A stable rank filtration on direct sum $K$-theory

Alexander Kupers; Inna Zakharevich; Jonathan Campbell

REVIEW 2 major objections 4 minor 25 references

A stable rank filtration on direct sum $K$-theory

T0 review · 2 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper builds a stable rank filtration on the K-theory spectrum of convenient addition categories, identifying filtration quotients as homotopy coinvariants and subobject K-theory as a suspension spectrum of decompositions.

desk verdict A promising filtration framework whose central computational theorem is not proved: the connectivity hypotheses in Corollary 1.19/1.20 fail for the configuration spaces that Proposition 3.13 produces. read the letter →

arxiv 2501.01609 v1 pith:LVUPWA2F submitted 2025-01-03 math.KT

Jonathan Campbell , Alexander Kupers , Inna Zakharevich This is my paper · ORCID

classification math.KT MSC 19D1019D2355P42

keywords algebraicK-theorystablerankfiltrationΓ-spacesconvenientadditioncategoriesdecompositionposetsspectralsequencescommonbasiscomplexBarratt–Priddy–Quillentheorem

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper's goal is a stable rank filtration on algebraic K-theory for a general class of symmetric monoidal categories, not just rings. It introduces convenient addition categories—categories in which direct sum behaves like a disjoint union—and shows that any rank-like valuation on such a category filters the K-theory spectrum at the spectrum level. The central result identifies the filtration quotients with homotopy coinvariants of the K-theory of subobject structures, and then identifies that subobject K-theory with the suspension spectrum of the nerve of the poset of nontrivial decompositions. From this it derives new spectral sequences converging to the homology of algebraic K-theory, recovers the Barratt–Priddy–Quillen theorem, and re-expresses Rognes's common basis complex.

What carries the argument

The load-bearing mechanism is the P-valuation: a functor from the category of elements of a Set-valued functor (or, levelwise, of a symmetric spectrum or Γ-set) to a poset P, which records the 'rank' of each element and induces a filtration F_p. The key identity is Theorem C, K(S_A) ≃ Σ^∞ $Σ^{1}$ N(Dcp^∘_A), where Dcp^∘_A is the poset of nontrivial unordered decompositions of A into non-initial subobjects, ordered by refinement. It is proven by Proposition 3.13, which shows the filtered pieces of the Γ-set S_A are exactly smash powers $L^{{∧|p|}}$, and by Corollary 1.19, which assembles these highly connected pieces into a stable equivalence via a levelwise Mayer–Vietoris argument. The axioms of a convenient addition category—unit is initial, all morphisms monic, maps out of a sum determined by components, and the pullback of the two inclusions into a sum is the initial object—are precisely what makes the decomposition poset well-behaved and the smash-power identification hold.

What would settle it

For A the category of finite-dimensional F_2-vector spaces and object A = $F_2^{3}$, Proposition 4.5 says N Dcp^∘_A is a wedge of circles, so the central equivalence predicts π_2(K(S_A)) ≅ H_1(N Dcp^∘_A; Z), a free abelian group of rank equal to the number of circles. Compute that rank from the poset of direct-sum decompositions of $F_2^{3}$ and independently compute π_2 of the Γ-set S_A from its simplicial sets S_A(S^k) for k = 1,2,3; any disagreement between the two groups would refute the central identification.

Watch

Extended reading notes

Core claim

For a convenient addition category A with a rank-like valuation, the K-theory spectrum K(A) carries a filtration indexed by the rank poset whose associated graded pieces are computed in two steps (Theorems 3.6 and 3.14). The first step is an Aut(A)-equivariant equivalence between the p-th filtration quotient and the homotopy coinvariants K(S_A)_{hAut(A)}, where S_A is the Γ-set built from the subobject structure of an object A of rank p. The second step is a general connectivity theorem (Corollary 1.19) applied to a decomposition-valued filtration on S_A: because the filtered pieces are highly connected smash powers, K(S_A) is stably equivalent to Σ^∞ $Σ^{1}$ N(Dcp^∘_A), the suspension spectrum of the nerve of the poset of nontrivial decompositions of A. In the applications, these decomposition posets are wedges of spheres (using Kupers's comparison with ordered decompositions and classical Tits-building results), which makes the spectral sequences collapse enough to compute rational K-groups from group homology.

Load-bearing premise

The argument that a filtered spectrum is a suspension spectrum of a nerve depends on the connectivity of the filtered pieces: at level n, every piece in the distinguished cosieve must be (2n−1)-connected, since only then does the levelwise approximation become a true stable equivalence; without that, the identification of K(S_A) and the spectral sequences built on it collapse.

Editorial extensions

If this is right

For finite sets, the filtration quotients above rank one vanish, giving a new proof of the Barratt–Priddy–Quillen theorem: K(FinSet_*) is the sphere spectrum.
For a field or Dedekind domain k, there is a spectral sequence with E1-page H_s(GL_n(k); Dec_n) converging to the rationalized K-groups of k, where Dec_n is the rational homology of the decomposition poset of k^n.
For inner product spaces over an ordered field, an analogous spectral sequence with orthogonal groups O_n(k) converges to the rational K-theory of that category.
Rognes's common basis complex C_n is shown to be stably equivalent to the suspension of the poset of nontrivial minimal spanning posets of R^n, and the paper conjectures this is an equivariant equivalence and that the poset is a wedge of (2n−3)-spheres.
The valuation formalism extends to Waldhausen categories, giving an alternate spectrum-level model of Rognes's rank filtration.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The suspension-spectrum form of K(S_A) suggests the whole rank filtration is cellular, so one could attempt to compute Steenrod operations or higher differentials in algebraic K-theory directly from the combinatorics of decomposition posets; the paper does not pursue this.
If the wedge-of-spheres conjecture for the minimal-spanning-poset space holds, the new spectral sequences would collapse at E2, yielding rational K-groups as a direct sum of group homologies of GL_n(k) (or O_n(k)) with Steinberg-like coefficients.
The recognition principle behind Theorem C—highly connected poset filtrations force a spectrum to be a suspension spectrum of a nerve—is stated for arbitrary spectra and could be applied to other filtered spectra in algebraic topology, such as Waldhausen's A-theory or L-theory, whenever a rank-like invariant exists.
The Aut(A)-equivariance in Theorem C provides extra structure not needed for the homology computations; tracking this action might yield homological stability theorems for automorphism groups of objects in convenient addition categories, generalizing stability for GL_n and symmetric groups.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Here's my take. The paper has a genuinely useful new framework—P-valuations on Γ-spaces and convenient addition categories—and Theorem 3.6, identifying filtration quotients as homotopy coinvariants of K(S_A), is a solid and attractive result. The writing is clear about the debts to Rognes and the GKRW circle, and the citation practice looks fine. But the main computational theorem, Theorem C, is not proved as written. The stress-test note is right. Proposition 3.13 identifies the filtered piece F_p S_A(L) with L^{|p|}, the pointed maps from |p| into L that are injective away from the basepoint. For L = S^n, that is the ordered configuration space Conf_{|p|}(S^n). For |p| ≥ 2, Conf_{|p|}(S^n) is at best (n−2)-connected—for example Conf_2(S^2) has π_1 = Z/2, and Conf_2(S^1) ≃ S^1. Corollary 1.19 needs the n-th filtered piece to be (2n−1)-connected, and Corollary 1.20 needs (2m+1)-connected for any m-connected L. The configuration spaces are nowhere near that connected. So the appeal to Corollaries 1.19/1.20 in Corollary 3.14 fails. That is load-bearing: the spectral sequences in Section 4 and the BPQ reproof for n>1 depend on Corollary 3.14. I do not see a quick patch in the text; the proof would need different connectivity bounds or a different identification of the filtered pieces. The theorem itself might be true, but it is unsupported here.

Smaller issues: the abstract promises convergence to the homology of algebraic K-theory, while the body only has rationalized statements; and the spectrum-level half of Theorem 3.6 is passed over quickly. These are minor.

What is good and likely durable: the P-valuation formalism, the convenient addition category setup, the decomposition poset analysis, and Kupers's appendix relating ordered and unordered decomposition posets. Those deserve to be in the literature.

Who should read this: people working on rank filtrations in K-theory, and anyone using Rognes's common basis complex. This is not a desk-reject; it is a paper with a serious gap in its central application. I would send it to a knowledgeable referee, with explicit instructions to check the connectivity hypothesis and to ask whether Corollary 3.14 can be replaced by a statement that holds. If the authors can fix that, the paper would be a real contribution.

Referee Report

2 major / 4 minor

Summary. The paper introduces a stable rank filtration on the direct-sum K-theory of 'convenient addition categories', built from P-valuations on Γ-spaces. It proves that the associated graded pieces of the filtration are homotopy coinvariants of the K-theory of a subobject structure S_A (Theorem A / Theorem 3.6), and it identifies the K-theory spectrum of S_A with the suspension spectrum of the nerve of the decomposition poset Dcp^∘_A (Theorem C / Corollary 3.14). This is then used to give new proofs and spectral sequences: the Barratt–Priddy–Quillen theorem, a rational spectral sequence for inner product spaces over ordered fields, a rational spectral sequence for free modules over fields or Dedekind domains, and a comparison with Rognes's common basis complex. The appendix by Kupers compares the homotopy type of decomposition posets with their ordered variants, following Mirzaii–van der Kallen.

Significance. If the main theorems hold, the paper gives a clean categorical framework that unifies and generalizes Rognes's stable rank filtration and common basis complex, and it supplies concrete spectral-sequence applications. The paper is well structured, contains substantial detailed proofs, and the appendix by Kupers is a useful contribution in its own right. The central computational engine, however, is Theorem C, and its proof currently relies on a connectivity hypothesis that is not satisfied; the issue is real but appears repairable by weakening the connectivity bounds, since only unbounded growth of the connectivity at spectrum level n is needed for a stable equivalence.

major comments (2)

[Corollary 1.20, Proposition 3.13, Corollary 3.14] The proof of Corollary 3.14 asserts that S_A satisfies the hypotheses of Corollary 1.20, but this is false. By Proposition 3.13, the filtered piece F_p S_A(L) is isomorphic to L^p, the pointed simplicial set of maps p → L that are constant or injective. For L = S^n and a nontrivial decomposition p with |p| ≥ 2, this is the ordered configuration space Conf_{|p|}(S^n) (up to the basepoint). Already for |p| = 2, Conf_2(S^n) is homotopy equivalent to S^n, which has connectivity n−1, not the 2n−1 required by Corollary 1.20 when applied to L = S^n (which is (n−1)-connected and forces m = n−1). For n = 2 this requires 3-connectivity, while Conf_2(S^2) ≃ S^2 is only 1-connected. Thus the hypotheses of Corollary 1.20 fail for every n ≥ 1. The same failure affects Corollary 1.22 and therefore the equivariant statements used in Sections 4.2 and 4.3. The gap is repairable: for a stable equivalence it is enough that the level-n map induce isomorphisms on π_i for i < n−2 (equivalently, for i < m in the notation of Corollary 1.20), because for fixed i this holds for all sufficiently large n. I recommend weakening the connectivity hypotheses in Corollaries 1.19–1.22 accordingly and rechecking the proof of Corollary 3.14.
[Sections 4.2–4.3, Corollary 4.4] The computations in Sections 4.2 and 4.3 depend on the equivariant equivalences K(S^n_k) ≃ Σ^∞ Σ^1 N Dcp^∘_{k^n} supplied by Corollary 3.14 via Corollary 1.22. Because the hypotheses of Corollary 1.22 are not satisfied, the displayed spectral sequences and the group identifications in Corollary 4.4 are currently unsupported. If the connectivity hypotheses are weakened as suggested in the previous comment, these applications should be verified levelwise; as written, the applications do not follow from the results proved in the paper.

minor comments (4)

[Abstract and Section 4] The abstract and introduction state that the spectral sequences converge to the homology of algebraic K-theory, but the body of the paper only proves convergence to rationalized K-groups. Please add the rational qualifier throughout the introductory statements.
[Lemma 1.18 proof] In the proof of Lemma 1.18, the word 'isomoprhism' should read 'isomorphism'.
[Section 4.3] The sentence about 'groups in green boxes being torsion' refers to colors that are absent in a monochrome printout; please use a color-independent designation or a displayed annotation.
[Example 3.9] The phrase 'square-root closed ordered field' is not defined; please define it or give a reference, since it is a condition on the field in an example.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation: Theorems A-C are proved from definitions and standard homotopy-theoretic tools; the connectivity gap in Corollary 3.14 is a correctness issue, not a circular one.

full rationale

I walked the derivation chain from Section 1 through Section 3 and the applications. Corollary 1.19 is proved directly from the Mayer-Vietoris blowup (Lemma 1.10), Lemma 1.18, and the standard localization result Theorem 1.17; it does not assume the suspension-spectrum conclusion. Theorem 3.6 is proved by an explicit comparison of the filtered Gamma-sets with K(S_A)_{hAut(A)} (Lemma 3.5 and the functor D), not by citing the conclusion. Proposition 3.13 gives a direct combinatorial identification F_p S_A(L) is isomorphic to L^p, and Corollary 3.14 invokes Corollary 1.20 as a recognition principle. The assertion in Corollary 3.14 that 'S_A satisfies the conditions of Corollary 1.20' is not accompanied by a verification of the 2m+1-connectivity of L^p for nontrivial p; that is an omitted proof or a potential false hypothesis (a correctness risk), not a circular reduction, because nowhere is K(S_A) equivalent to Sigma^infinity Sigma^1 N Dcp^circ_A fed back into the hypotheses. Applications invoke external theorems (Quillen, Charney, Mirzaii-van der Kallen, and GKRW for a vanishing line) as tools; these are not used to derive the central equivalence. Self-references to Rognes's common basis complex and to GKRW are motivational or contextual or support peripheral vanishing claims, and no obtained result is renamed as a new prediction. No fitted parameter is renamed as a prediction, and no uniqueness theorem from the authors' own prior work is used to force a choice. Therefore no step of the claimed derivation reduces by construction to its own input.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central theorems rest on the convenient addition category axioms, the valuation reflecting isomorphisms, connectivity hypotheses, and the rank-chain length assumption. No free parameters or invented physical entities appear; the paper is a purely mathematical structural result.

assumptions (5)

domain assumption Convenient addition category axioms (CA1)-(CA3): ∅ is initial, all morphisms monic, maps out of A⊕B are determined by components, and A ×_{A⊕B} B = ∅.
The paper restricts attention to categories satisfying these conditions; examples are claimed to satisfy them, but this is a hypothesis on the input categories.
domain assumption The valuation ν: A_mon → P reflects isomorphisms: if ν(A)=ν(B) then A≅B.
Required in Theorem 3.6 to identify the filtration quotient with the homotopy coinvariants K(S_A)_{hAut(A)}; without this, the right-hand side is not well-defined.
domain assumption Connectivity hypothesis in Corollary 1.19: for each n, if A∈P^∘ then F_A X_n is (2n−1)-connected.
This is the technical input that turns a levelwise approximation into a stable equivalence; it is assumed, not derived.
domain assumption Rank axiom of Appendix A: A has a rank function such that the longest chain in Dcp(A) has length rk(A)−1.
Used in Theorem A.4 to compare the decomposition poset to the ordered decomposition poset; it holds in the examples but is a restriction.
standard math Standard results from homotopy theory, including left Bousfield localization, Farjoun's homotopy colimit theorem, Mirzaii-van der Kallen connectivity lemma, and sphericity of Tits buildings.
The proofs rely on these external results, cited and used without proof.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A stable rank filtration on direct sum $K$-theory." pith.science (2026). https://pith.science/paper/LVUPWA2F

@misc{pith2026250101609,
  author       = {Pith},
  title        = {Pith review of: A stable rank filtration on direct sum $K$-theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LVUPWA2F}},
  note         = {Machine review of arXiv:2501.01609}
}

abstract

In the literature, there are two standard rank filtrations on $K$-theory: an ``unstable'' one which is traditionally defined through the homology of $GL_n$, and a ``stable'' one which was defined by Rognes using the simplicial structure on Waldhausen's $S_\bullet$-construction. In this paper we give an alternate stable rank filtration, which uses the simplicial structure present in a $\Gamma$-space construction of $K$-theory; we investigate this in the case of ``convenient addition categories,'' and show that in good situtations where a notion of ``rank'' is present, the filtration quotients will be homotopy coinvariants of certain highly-connected suspension spectra. This approach generalizes Rognes's results on the common basis complex, and produces an alternate spectral sequences converging to the homology of algebraic $K$-theory.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 22 canonical work pages

[1]

On (enriched) left B ousfield localization of model categories, 2007

Clark Barwick. On (enriched) left B ousfield localization of model categories, 2007

work page 2007
[2]

Handbook of Categorical Algebra

Francis Borceux. Handbook of Categorical Algebra . Encyclopedia of Mathematics and its Applications. Cambridge University Press, 1994

work page 1994
[3]

On the homology of non-connected monoids and their associated groups

Michael Barratt and Stewart Priddy. On the homology of non-connected monoids and their associated groups. Comment. Math. Helv. , 47:1--14, 1972

work page 1972
[4]

Ruth M. Charney. Homology stability for gln of a dedekind domain. Inventiones mathematicae , 56:1--18, 1980

work page 1980
[5]

Cohomology of braid spaces

Fred Cohen. Cohomology of braid spaces . Bulletin of the American Mathematical Society , 79(4):763 -- 766, 1973

work page 1973
[6]

Jens Niklas Eberhardt, Oliver Lorscheid, and Matthew B. Young. Group completion in the k-theory and grothendieck–witt theory of proto-exact categories. Journal of Pure and Applied Algebra , 226(8):107018, 2022

work page 2022
[7]

A. D. Elmendorf and M. A. Mandell. Rings, modules, and algebras in infinite loop space theory. Adv. Math. , 205(1):163--228, 2006

work page 2006
[8]

Cellular spaces, null spaces and homotopy localization , volume 1622 of Lecture Notes in Mathematics

Emmanuel Dror Farjoun. Cellular spaces, null spaces and homotopy localization , volume 1622 of Lecture Notes in Mathematics . Springer-Verlag, Berlin, 1996

work page 1996

Show all 25 references

[9]

Iterated bar complexes of E –infinity algebras and homology theories

Benoit Fresse. Iterated bar complexes of E –infinity algebras and homology theories . Algebraic & Geometric Topology , 11(2):747 -- 838, 2011

2011
[10]

Galatius, A

S. Galatius, A. Kupers, and O. Randal-Williams. Cellular E_k -algebras. to appear in Ast\'erisque, https://arxiv.org/abs/1805.07184, 2018

2018 arXiv
[11]

Galatius, A

S. Galatius, A. Kupers, and O. Randal-Williams. E _ -cells and general linear groups of infinite fields. https://arxiv.org/abs/2005.05620, 2020

2005 arXiv
[12]

Configuration spaces in algebraic topology, 2018

Ben Knudsen. Configuration spaces in algebraic topology, 2018. arXiv:1803.11165

2018 arXiv
[13]

Kadeishvili and S

T. Kadeishvili and S. Saneblidze. Iterating the bar construction. Georgian Math. J. , 5(5):441--452, 1998

1998
[14]

Jeremy Miller, Peter Patzt, and Jennifer C. H. Wilson. On rank filtrations of algebraic k-theory and steinberg modules, 2023

2023
[15]

Mirzaii and W

B. Mirzaii and W. van der Kallen. Homology stability for unitary groups. Doc. Math. , 7:143--166, 2002

2002
[16]

Finite generation of the groups ki of rings of algebraic integers

Daniel Quillen. Finite generation of the groups ki of rings of algebraic integers. In H. Bass, editor, Higher K-Theories , pages 179--198, Berlin, Heidelberg, 1973. Springer Berlin Heidelberg

1973
[17]

J. Rognes. A spectrum level rank filtration in algebraic K -theory. Topology , 31(4):813--845, 1992

1992
[18]

The weight and rank filtrations

John Rognes. The weight and rank filtrations. arXiv: 2110.12264, 2021

2021 arXiv
[19]

Randal-Williams

O. Randal-Williams. What is the homotopy type of the poset of nontrivial decompositions of R^n ? MathOverflow, 2022. https://mathoverflow.net/q/414964

2022
[20]

Symmetric spectra

Stefan Schwede. Symmetric spectra. http://www.math.uni-bonn.de/people/schwede/SymSpec-v3.pdf
[21]

Categories and cohomology theories

Graeme Segal. Categories and cohomology theories. Topology , 13(3):293--312, 1974

1974
[22]

Spherical complexes attached to symplectic lattices

Wilberd van der Kallen and Eduard Looijenga. Spherical complexes attached to symplectic lattices. Geom. Dedicata , 152:197--211, 2011

2011
[23]

Waldhausen

F. Waldhausen. Algebraic K -theory of spaces. In Algebraic Geometry and Topology (New Brunswick, N.J, 1983) , volume 1126 of Lecture Notes in Mathematics . Springer, 1985

1983
[24]

Charles A. Weibel. The K -book , volume 145 of Graduate Studies in Mathematics . American Mathematical Society, Providence, RI, 2013. An introduction to algebraic K -theory

2013
[25]

V. Welker. Direct sum decompositions of matroids and exponential structures. Journal of Combinatorial Theory, Series B , 63(2):222--244, 1995

1995

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [1]

On (enriched) left B ousfield localization of model categories, 2007

Clark Barwick. On (enriched) left B ousfield localization of model categories, 2007

work page 2007

[2] [2]

Handbook of Categorical Algebra

Francis Borceux. Handbook of Categorical Algebra . Encyclopedia of Mathematics and its Applications. Cambridge University Press, 1994

work page 1994

[3] [3]

On the homology of non-connected monoids and their associated groups

Michael Barratt and Stewart Priddy. On the homology of non-connected monoids and their associated groups. Comment. Math. Helv. , 47:1--14, 1972

work page 1972

[4] [4]

Ruth M. Charney. Homology stability for gln of a dedekind domain. Inventiones mathematicae , 56:1--18, 1980

work page 1980

[5] [5]

Cohomology of braid spaces

Fred Cohen. Cohomology of braid spaces . Bulletin of the American Mathematical Society , 79(4):763 -- 766, 1973

work page 1973

[6] [6]

Jens Niklas Eberhardt, Oliver Lorscheid, and Matthew B. Young. Group completion in the k-theory and grothendieck–witt theory of proto-exact categories. Journal of Pure and Applied Algebra , 226(8):107018, 2022

work page 2022

[7] [7]

A. D. Elmendorf and M. A. Mandell. Rings, modules, and algebras in infinite loop space theory. Adv. Math. , 205(1):163--228, 2006

work page 2006

[8] [8]

Cellular spaces, null spaces and homotopy localization , volume 1622 of Lecture Notes in Mathematics

Emmanuel Dror Farjoun. Cellular spaces, null spaces and homotopy localization , volume 1622 of Lecture Notes in Mathematics . Springer-Verlag, Berlin, 1996

work page 1996