Categorical foundations of discrete dynamical systems

Chris Kapulkin; Daniel Carranza; Matthew Wheeler; Nathan Kershaw; Reinhard Laubenbacher

REVIEW 1 major objections 6 minor 8 references

Categorical foundations of discrete dynamical systems

T0 review · 1 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Attractors of a semi-direct product of discrete dynamical systems decompose into a disjoint union of attractors of smaller twisted component systems, generalizing the Boolean-network modularity theorem to all discrete dynamical systems.

desk verdict Genuinely new categorical framework for attractors of discrete dynamical systems; the decomposition theorem is the real result, and it checks out. read the letter →

arxiv 2506.05190 v1 pith:U4CUJHPK submitted 2025-06-05 math.DS math.CT

Daniel Carranza , Chris Kapulkin , Nathan Kershaw , Reinhard Laubenbacher , Matthew Wheeler This is my paper

classification math.DSmath.CT MSC 37B2518A3518A4005C2037B10

keywords discretedynamicalsystemscategorytheorycyclesetsattractorssemi-directproductBooleannetworkswiringdiagrammodularity

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

Discrete dynamical systems—Boolean networks, Petri nets, cellular automata, and any set with an update function—are usually analyzed case by case. This paper asks whether the structure of such a system determines its attractors in a uniform way, and answers with a categorical framework built around a new object called a cycle set. The load-bearing result says that if a system is a semi-direct product of two smaller systems, then its attractors are exactly a disjoint union of attractors of the smaller systems, with one fiber system per orbit of attractors of the base. That recovers a known Boolean-network modularity theorem as a special case and extends it to arbitrary discrete dynamical systems, including ones whose state spaces are infinite. The payoff would be a structural, decomposition-based way to analyze dynamics across all these model families at once.

What carries the argument

The central object is the cycle set: a collection of sets $K_n$, one for each cycle length $n$, with a $\mathbb{Z}/n$-rotation action and degeneracy maps that repeat an $n$-cycle to an $mk$-cycle, organized as a functor from the category of directed cycles to sets. It is the formal language in which attractors are spoken: the attractor functor $A$ sends a digraph or state space to its cycle set. The proof of the decomposition theorem runs through the fact that the state space functor, the attractor functor, and the evaluation functor $\mathrm{ev}_n$ all preserve limits, so the pullback square identifying a semi-direct product is preserved after taking attractors; the index set then decomposes by orbit-stabilizer over the base's cycle orbits. The wiring diagram of a map $f:A^n\to A^n$ is the companion mechanism used to detect semi-direct products: $f$ is a semi-direct product, up to permuting coordinates, exactly when its wiring diagram maps to the two-loop one-edge digraph $E^\circ$.

What would settle it

Compute both sides of the theorem explicitly for a small semi-direct product, say the Boolean network $f(x_1,x_2)=(x_1+x_2,x_1)$ over $\mathbb{F}_2$ from the paper's Example 2.13: enumerate the cycles of all lengths of the four-state system and compare them with the disjoint union over the base system's cycle orbits of the cycle sets of the twisted fiber systems. Any mismatch in the number of $n$-cycles, or in the $\mathbb{Z}/n$-orbit structure, for any $n$ would refute Theorem 2.3. A softer check targets the stated convention: drop the degenerate-cycle multiplicities from the definition of a cycle set and repeat the pullback computation; the preservation-of-limits step should fail at a small example such as a single 2-cycle base, showing exactly where the theorem depends on that convention.

Watch

Extended reading notes

Core claim

The paper's central claim is that the passage from a discrete dynamical system to its attractors is a categorical construction that respects decomposition. Concretely, for any semi-direct product $(X\times Y, f \rtimes_p g)$, there is an isomorphism of $\mathbb{Z}/n$-sets $$AS(X\times Y, f\rtimes_p g)_n \cong \coprod_{[c]\in \mathrm{Orb}(AS(X,f)_n)} AS(\mathbb{Z}/k \times Y, \mathrm{succ}\rtimes_{p_c} g)_n,$$ where $k$ is the length of the non-degenerate cycle $c$ underlying the orbit $[c]$, and $p_c$ is the twist along that cycle. In words: every $n$-cycle of the combined system lies in exactly one component indexed by an orbit of $n$-cycles of the base, and within that component it is an $n$-cycle of a smaller system built from the attractor cycle and the second factor. The theorem is obtained by showing that the whole pipeline from dynamical systems to digraphs to cycle sets to $\mathbb{Z}/n$-sets preserves limits, so the pullback square expressing the semi-direct product is still a pullback after taking attractors.

Load-bearing premise

The load-bearing convention is that cycles are stored with a chosen ordering and every $n$-cycle also generates degenerate $mk$-cycles for every multiple $m$; without this extra data, the cycle-set construction would not preserve the limits used in the proof, and the decomposition theorem would not go through.

Editorial extensions

If this is right

Every semi-direct product decomposition of a discrete dynamical system now yields a computational reduction: attractors of the whole are built from attractors of the base and of smaller twisted fiber systems.
Because the functors preserve limits and coproducts, products and disjoint unions of systems can be analyzed by performing the corresponding constructions on cycle sets.
The wiring-diagram criterion gives a structural certificate: a system admits a semi-direct product decomposition iff its wiring diagram maps to $E^\circ$, so decomposability can be read off a graph.
The framework applies beyond finite sets: the same results hold in any category with finite products, including topological spaces and smooth manifolds.
Cycle sets satisfying the paper's Properties A and B are exactly those arising from digraphs, so non-degenerate attractor counts can be recovered from orbit data.
The self-referential remark in the introduction is correct that the preservation results would fail without the conventions that cycles are ordered and that every $n$-cycle generates degenerate $mk$-cycles; those conventions are part of Definition 1.20 and are relied on in the proof of Theorem 2.3.

Reading between the lines

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

A categorical definition of modularity as 'the attractor cycle set is a semi-direct product of component cycle sets' would make the biological notion of modularity independent of any particular network model; the paper does not state this definition explicitly.
The wiring-diagram criterion suggests a search procedure: for any fixed finite target digraph $T$, systems whose wiring diagram maps to $T$ should form a class with a corresponding decomposition law, generalizing the $E^\circ$ case; the paper mentions this as future work, and the algorithmic reading is our inference.
Because full cycle sets contain infinite degenerate data, any implementation would need to truncate at a maximum cycle length; bounding the error of such truncations is an open engineering question not addressed in the paper.
If the same presheaf construction is attempted for continuous time, replacing $\mathbb{N}$ by $\mathbb{R}$ and cycles by orbits of a circle action, a decomposition theorem would have to sum or integrate over period types; the paper only gestures at this hoped-for generalization.

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

Read it; this one deserves referee time. Carranza et al. build a category-theoretic language for discrete dynamical systems where attractors become cycle sets, and they prove that semi-direct products decompose at the level of attractors. The decomposition theorem (2.3) generalizes the Boolean-network result from Kadelka et al. 2023 to arbitrary DDS, and the wiring-diagram characterization (2.15) gives a clean structural test for semi-direct product structure. Those two results are new, not just re-proved special cases.

The paper is honest about its main design choice: cycles are stored with an ordering and degenerate multiples are included, so the cycle set functor preserves limits. The authors state plainly that the preservation results would fail without this. That is a real restriction on what counts as an 'attractor,' but it is a deliberate convention, and passing to non-degenerate orbits recovers the usual objects. It is not a hidden flaw.

I checked the proof of Theorem 2.3 with some care. The composite DDS -> DiGraph -> cySet -> Set(Z/n) preserves limits by Cor 1.14, Prop 1.25, and Cor 1.22. Proposition 2.2 gives a genuine pullback square, and the pullback argument in Set(Z/n) is locally Cartesian closed, so the orbit-wise decomposition is sound. The 'moreover' independence claim also holds: rotating the chosen representative changes p_c by a rotation of the cycle, and Prop 2.2 supplies the iso. The proof is intricate; a machine-checked version would be welcome, but I did not find a gap that matters.

The soft spots are minor. Theorem 1.32(2) and Prop 1.37 compress a uniqueness argument; the reader has to fill in a gcd/stabilizer detail. It is fillable. The paper claims accessibility for non-category-theorists but the body requires working knowledge of functor categories and adjoints; the intro is a nice summary, not a tutorial. And as a methodology paper it carries no new computational examples, so applied readers will want to see the framework used on a real network before adopting it.

Who is it for? Researchers in discrete dynamics, especially Boolean network modularity, and category theorists looking for applied targets. It should go to peer review, not desk reject.

Referee Report

1 major / 6 minor

Summary. This paper develops a categorical framework for discrete dynamical systems. The category DDS of discrete dynamical systems is defined as presheaves on the monoid N, and the state space functor into digraphs is shown to be full and faithful. The central new object is the category cySet of cycle sets, defined as presheaves on the category of directed cycles C•; a cycle set stores n-cycles with an ordering and with all degenerate mk-cycles for each multiple. The attractor functor A: DiGraph→cySet is shown to preserve limits and coproducts (Propositions 1.25 and 1.26), and a recognition theorem (Theorem 1.40) characterizes cycle sets arising from digraphs via Properties A and B. As a proof of concept, the paper proves a decomposition theorem (Theorem 2.3): for a semi-direct product f ⋊_p g, the n-cycles of the attractors decompose as a disjoint union over orbits of n-cycles of the factor systems of the attractors of twisted component systems, generalizing the Boolean-network modularity theorem of Kadelka et al. [KWVC+23]. The paper also generalizes wiring diagrams to arbitrary categories with finite products and characterizes semi-direct product morphisms via graph maps to the walking looped-edge graph E◦ (Theorem 2.15).

Significance. The paper's significance lies in providing a common categorical language for attractors of Boolean networks, Petri nets, cellular automata, and other discrete dynamical systems, and in showing that a modularity decomposition theorem can be derived from abstract preservation properties. The proof of Theorem 2.3 is explicit and relies on proved limit preservation results rather than imported machinery; the 'moreover' independence of representatives is also justified via rotation isomorphisms. The recognition theorem and the wiring diagram characterization are new and potentially useful. The authors are candid about the conventions (ordered and degenerate cycles) that make the limit-preservation results possible, and about the admissibility conditions in the wiring diagram section. These design choices are stated clearly and are not hidden assumptions. Overall, if the central claims hold, this is a valuable foundation for future work on modularity and decomposition of discrete dynamical systems.

major comments (1)

[Theorem 1.32(2) and Proposition 1.37] The proofs as written establish only the uniqueness of the non-degenerate representative, not its existence. This is a load-bearing point because the statement of Theorem 2.3 chooses, for each orbit [c], an element c~ = c μ_{n,k} with c non-degenerate, which presupposes existence. The missing existence argument is a standard minimal-length selection: among all m for which the given cycle is a degeneracy of an m-cycle, choose the least; the resulting m-cycle is then non-degenerate by minimality. Please add this argument in both places, and verify that the digraph case (Theorem 1.32) does not require Property B for existence.

minor comments (6)

[Definition 2.12] The word 'veritices' should be 'vertices'.
[Example 1.34] The word 'satisfes' should be 'satisfies'.
[Lemma 1.39] The phrase 'we ˆC_k for the cycle set' is missing a verb; it should read 'we write \hat{C}_k for the cycle set'.
[Lemma 2.9] In the statement, the projection should be typed as proj_{≠i_1,...,i_k} : A^{n+k} → A^n, not A^n → B; the diagram should use the same notation.
[Theorem 2.3] The map p_c used in the statement is not defined; it should be defined as the composite p∘c : Z/k → E (or an explicit note that p_c = p∘c).
[Proof of Theorem 2.3] The assertion that AS(c)_n has image exactly the orbit [c] and is an isomorphism onto it is correct but terse; a short justification using non-degeneracy of c and Theorem 1.32(1) would aid readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Theorem 2.3 is derived from internal definitions and proved preservation results, not from a fitted parameter or a load-bearing self-citation.

full rationale

The central result, Theorem 2.3, is obtained by a chain of internally established facts: the composite DDS → DiGraph → cySet → SetZ/n preserves limits (Corollary 1.14, Proposition 1.25, Corollary 1.22), the square from Proposition 2.2 is a genuine pullback, and SetZ/n being locally Cartesian closed converts the orbit decomposition into the asserted level-wise attractor decomposition. The cited Boolean-network decomposition theorem [KWVC+23] is used only as motivation and as the special case being generalized; it is not assumed as an input to the proof. The paper's explicit conventions about ordered cycles and degenerate cycles are design choices, openly stated as necessary for limit preservation, and they do not embed the conclusion of Theorem 2.3. No fitted parameters, predictions, or uniqueness theorems imported from the authors' prior work appear in the derivation. The self-citation to [KWVC+23] is therefore not load-bearing, and the derivation is self-contained.

Assumptions & free parameters 0 free parameters · 5 assumptions · 2 invented entities

The central claim rests on standard set-theoretic and categorical machinery, plus two conventions introduced by the paper: ordered cycles with degenerate repetitions, and admissible products. Neither convention has an external empirical status; they are design choices that make the preservation theorems and the wiring-diagram characterization work. The paper introduces two new mathematical objects, cycle sets and generalized semi-direct products, whose value is verified only through the theorems proved here.

assumptions (5)

standard math ZFC set theory with standard category theory: presheaf categories, limits, colimits, adjoint functors, and the orbit-stabilizer theorem.
Used throughout Sections 1 and 2 without proof; e.g., Definition 1.1, Proposition 1.10, Theorem 1.32.
domain assumption A discrete dynamical system is modeled as a set with an endofunction, equivalently a functor BN→Set; morphisms are equivariant maps.
Definition 1.1. This is the paper's chosen meaning of discrete dynamical system and is standard.
ad hoc to paper Cycles are recorded with a chosen ordering, and each n-cycle generates all degenerate mk-cycles; the indexing category C• consists of directed cycle graphs with wrap maps μ and rotations ρ.
Definition 1.20 and the discussion before it. The paper notes preservation theorems would fail without these conventions.
ad hoc to paper Products A^n must be admissible: if n=1, A must have a morphism from the terminal object.
Definition 2.6. This technical condition is needed for Proposition 2.7 and the wiring-diagram characterization in Theorem 2.15.
standard math Choice of a distinguished representative from each orbit in the proof of Theorem 2.3.
The theorem states let c~ be a distinguished element; this uses the axiom of choice when there are infinitely many orbits.

invented entities (2)

Cycle sets
purpose: Formal presheaf language for the attractors of a directed graph or dynamical system
Defined in Definition 1.20 as functors on the category of directed cycles. This is a new mathematical object introduced by the paper; its utility is internal to the framework and no external empirical handle is provided.
Semi-direct product of dynamical systems in arbitrary categories with finite products
purpose: Generalizes the Boolean-network semi-direct product from [KWVC+23] to any such category, enabling the decomposition theorem
Definition 2.1 is a new abstraction. It recovers the prior Boolean-network notion as a special case, but its broader value is not yet tested outside this paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Categorical foundations of discrete dynamical systems." pith.science (2026). https://pith.science/paper/U4CUJHPK

@misc{pith2026250605190,
  author       = {Pith},
  title        = {Pith review of: Categorical foundations of discrete dynamical systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/U4CUJHPK}},
  note         = {Machine review of arXiv:2506.05190}
}

read the original abstract

We develop categorical foundations of discrete dynamical systems, aimed at understanding how the structure of the system affects its dynamics. The key technical innovation is the notion of a cycle set, which provides a formal language in which to speak of the system's attractors. As a proof of concept, we provide a decomposition theorem for discrete dynamical systems.

Figures

Figures reproduced from arXiv: 2506.05190 by the authors.

**Figure 1.** The state space of a discrete dynamical system (in fact, a Boolean network) 𝑓 : {0, 1} 3 → {0, 1} 3 given by 𝑓 (𝑥1, 𝑥2, 𝑥3) = (𝑥1, 𝑥2 ⊕ 𝑥3, 𝑥1 ∨ 𝑥2). The goal of this paper is to change this paradigm by developing a foundational framework that permits an en-masse analysis. The cornerstone of this framework is the language of category theory, a branch of mathematics well suited for understanding abstract structure… view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

8 extracted references · 7 canonical work pages

[10]

Gan and R

MR3776769 [GA18] X. Gan and R. Albert, General method to find the attractors of discrete dynamic models of biological systems , Phys. Rev. E97 (2018), no. 4, 042308,

work page 2018
[11]

Gamache-Poirier, A

29 [GPSL+25] S. Gamache-Poirier, A. Souvane, W. Leclerc, C. Villeneuve, and S. V Hardy, An algorithm for the transformation of the petri net models of biological signaling networks into influence graphs , bioRxiv (2025), 2025–01. [HH24] S. Huitzil and C. Huepe, Life’s building blocks: the modular path to multiscale complexity , Frontiers in Systems Biolog...

arXiv 2025
[18]

Gardner, The fantastic combinations of john conway’s new solitaire game “life” , Scientific American 223 (1970), no

MR3807681 [Gar70] M. Gardner, The fantastic combinations of john conway’s new solitaire game “life” , Scientific American 223 (1970), no. 120-123,

work page 1970
[586]

[FSR16] B. Fong, P. Soboci ´nski, and P. Rapisarda, A categorical approach to open and interconnected dynamical systems , Proceedings of the 31st Annual ACM-IEEE Symposium on Logic in Computer Science (LICS 2016), 2016, pp

work page 2016
[2008]

[VCAHL14] A

Notes for Math 254A Ergodic theory. [VCAHL14] A. Veliz-Cuba, B. Aguilar, F. Hinkelmann, and R. Laubenbacher, Steady state analysis of boolean molecular network models via model reduction and computational algebra , BMC bioinformatics 15 (2014), 1–8. [WPC07] G. P. Wagner, M. Pavlicev, and J. M. Cheverud, The road to modularity, Nature Reviews Genetics 8 (D...

work page 2014
[2014]

Shilts, Y

MR3288752 [SSG+22] J. Shilts, Y. Severin, F. Galaway, N. M¨uller-Sienerth, Z.-S. Chong, S. Pritchard, S. Teichmann, R. Vento-Tormo, B. Snijder, and G. J. Wright, A physical wiring diagram for the human immune system , Nature 608 (August 2022), no. 7922, 397–404 (en). [SSV20] P. Schultz, D. I. Spivak, and C. Vasilakopoulou, Dynamical systems and sheaves , ...

work page 2020
[2015]

MR3347092 [CT18] M

Thesis (Ph.D.)–Rutgers The State University of New Jersey - New Brunswick. MR3347092 [CT18] M. Chaves and L. Tournier, Analysis tools for interconnected boolean networks with biological applications , Frontiers in physiology 9 (2018),

work page 2018
[2017]

[BKP+17] M

preprint. [BKP+17] M. Behrisch, S. Kerkhoff, R. P ¨oschel, F. M. Schneider, and S. Siegmund, Dynamical systems in categories , Appl. Categ. Structures 25 (2017), no. 1, 29–57. MR3606493 [Bus15] J. Bush, Shift equivalence and a combinatorial-topological approach to discrete-time dynamical systems , ProQuest LLC, Ann Arbor, MI,

work page 2017

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [10]

Gan and R

MR3776769 [GA18] X. Gan and R. Albert, General method to find the attractors of discrete dynamic models of biological systems , Phys. Rev. E97 (2018), no. 4, 042308,

work page 2018

[2] [11]

Gamache-Poirier, A

29 [GPSL+25] S. Gamache-Poirier, A. Souvane, W. Leclerc, C. Villeneuve, and S. V Hardy, An algorithm for the transformation of the petri net models of biological signaling networks into influence graphs , bioRxiv (2025), 2025–01. [HH24] S. Huitzil and C. Huepe, Life’s building blocks: the modular path to multiscale complexity , Frontiers in Systems Biolog...

arXiv 2025

[3] [18]

Gardner, The fantastic combinations of john conway’s new solitaire game “life” , Scientific American 223 (1970), no

MR3807681 [Gar70] M. Gardner, The fantastic combinations of john conway’s new solitaire game “life” , Scientific American 223 (1970), no. 120-123,

work page 1970

[4] [586]

[FSR16] B. Fong, P. Soboci ´nski, and P. Rapisarda, A categorical approach to open and interconnected dynamical systems , Proceedings of the 31st Annual ACM-IEEE Symposium on Logic in Computer Science (LICS 2016), 2016, pp

work page 2016

[5] [2008]

[VCAHL14] A

Notes for Math 254A Ergodic theory. [VCAHL14] A. Veliz-Cuba, B. Aguilar, F. Hinkelmann, and R. Laubenbacher, Steady state analysis of boolean molecular network models via model reduction and computational algebra , BMC bioinformatics 15 (2014), 1–8. [WPC07] G. P. Wagner, M. Pavlicev, and J. M. Cheverud, The road to modularity, Nature Reviews Genetics 8 (D...

work page 2014

[6] [2014]

Shilts, Y

MR3288752 [SSG+22] J. Shilts, Y. Severin, F. Galaway, N. M¨uller-Sienerth, Z.-S. Chong, S. Pritchard, S. Teichmann, R. Vento-Tormo, B. Snijder, and G. J. Wright, A physical wiring diagram for the human immune system , Nature 608 (August 2022), no. 7922, 397–404 (en). [SSV20] P. Schultz, D. I. Spivak, and C. Vasilakopoulou, Dynamical systems and sheaves , ...

work page 2020

[7] [2015]

MR3347092 [CT18] M

Thesis (Ph.D.)–Rutgers The State University of New Jersey - New Brunswick. MR3347092 [CT18] M. Chaves and L. Tournier, Analysis tools for interconnected boolean networks with biological applications , Frontiers in physiology 9 (2018),

work page 2018

[8] [2017]

[BKP+17] M

preprint. [BKP+17] M. Behrisch, S. Kerkhoff, R. P ¨oschel, F. M. Schneider, and S. Siegmund, Dynamical systems in categories , Appl. Categ. Structures 25 (2017), no. 1, 29–57. MR3606493 [Bus15] J. Bush, Shift equivalence and a combinatorial-topological approach to discrete-time dynamical systems , ProQuest LLC, Ann Arbor, MI,

work page 2017