REVIEW 4 major objections 6 minor 3 references
An explicit power series result for the two type ASEP
T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read This paper derives an explicit power-series generating function for the two-type ASEP with one type-2 particle, one type-1 particle, and any number of holes, expressed through one-type walk enumerators.
desk verdict A plausible formula for a two-type ASEP generating function, but the proof of the key decomposition is missing and the theorem statements contain internal inconsistencies. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the decomposition of a two-type ASEP walk into excursions between visits to the distinguished boundary point $(m,0)$, together with the lattice-walk bijection from [2] that represents ASEP states as walks in a simplicial region. The one-type walk generating functions of Theorems 3.1--3.3 are the building blocks: $G_{m,0,m,0}$ counts loops that return to $(m,0)$, $G_{m,0,0,m}$ counts crossings from $(m,0)$ to $(0,m)$, and $G_{m,0}$ and $G_{0,m}$ count final tails from either endpoint. The combinatorial step is to mark a forward exchange of the $1$ and $2$ particles by $d$, a reverse exchange by $1/d$, and to observe that after a forward exchange the walk must cross to $(0,m)$ before the reverse exchange can occur; the geometric series in $G_{m,0,0,m}^2$ then sums over all numbers of reverse-swap round trips.
What would settle it
For $m=1$, enumerate the six-state ASEP on $\{2,1,0\}$ by transfer matrix, using the paper's convention that $t$ counts steps: take the starting configuration with $1$ and $2$ adjacent, give each transition a $t$-weight per step, form the generating function by path length, and compare coefficients with the expansion of the right-hand side of Theorem 4.1 at $x=y=1$, $d=1$. Agreement through many terms supports the decomposition; the first disagreement would show that the $(m,0)$-return factorization does not count the same walks as the ASEP.
Extended reading notes
Core claim
The central claim is Theorem 4.1: for $\lambda=(2,1,0,\dots,0)$ with $m$ zeroes, $$A_\$\lambda$(x,y;t;d)=G_{m,0}(x,y;t)+\frac{G_{m,0,m,0}(1,1;t)\left(dG_{m,0}(x,y;t)+G_{m,0,0,m}(1,1;t)G_{0,m}(x,y;t)\right)}{1-G_{m,0,0,m}(1,1;t)^2}.$$ Here $G_{m,0}(x,y;t)$ counts walks in the linear lattice from $(m,0)$ to a final point recorded by $x$ and $y$; $G_{m,0,m,0}(1,1;t)$ counts walks that start and end at $(m,0)$; and $G_{m,0,0,m}(1,1;t)$ counts walks from $(m,0)$ to $(0,m)$. The formula is assembled by waiting at $(m,0)$ for the first exchange of the $1$ and $2$, then factoring all later exchanges into round trips from one distinguished endpoint to the other.
Load-bearing premise
The formula rests on the claim that, in this two-type ASEP, the $1$ and $2$ particles exchange positions only when the associated lattice walk is at the boundary point $(m,0)$, and that the walk pieces between such visits are independent and counted exactly by the one-type generating functions; the paper states this decomposition in Section 4 without proof.
Editorial extensions
If this is right
- Coefficient extraction in $t$, $x$, $y$, and $d$ from $A_\lambda$ gives exact counts of $n$-step two-type ASEP paths for the particle content $(2,1,0^m)$, with a specified final lattice point and a specified net number of exchanges of the two particle types.
- Setting $d=1$ recovers the unweighted two-type path generating function, while the coefficient of $d^k$ isolates paths whose net number of forward exchanges is $k$.
- Because every $G$ in the formula is ultimately an explicit function of $t$ and $p$ from Theorems 3.1--3.3, the identity is fully explicit: a reader can expand $A_\lambda$ to any desired order in $t$ without further combinatorial work.
- The form of the identity does not change as $m$ grows; increasing the number of holes only changes the indices of the one-type generating functions, so the two-type result scales in a uniform way.
Reading between the lines
- If the $(m,0)$-return decomposition is correct, a similar switch-marking argument should produce formulas for $\lambda=(2,1,0^m)$ with the two particles not initially adjacent, by first running the one-type walk until the particles meet; the paper states only the adjacent-start case.
- The $d$ variable records net exchanges, because a reverse exchange carries $1/d$ and cancels a forward $d$; a reader who wants the distribution by total number of swaps would need a separate variable, a modification not written in the paper.
- The paper's own future-work remark that higher simplices are needed suggests that for three or more particle types the single distinguished point $(m,0)$ would have to be replaced by a hierarchy of boundary faces, so the present formula is best read as a base case for a tree of excursion generating functions.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims an explicit power series for the two-type ASEP on a ring for the partition λ = (2,1,0,...,0) with m zeroes. The main result, Theorem 4.1, expresses the generating function Aλ(x,y;t;d) as a rational combination of generating functions G_{u,v} for walks in a simplified linear lattice, drawing on theorems of Mortimer and Prellberg and on Elizalde's bijection between ASEP configurations and walks in simplicial regions. The proof strategy is to decompose the path into segments between supposed 1/2 swap events and to count those segments using the one-type walk generating functions.
Significance. If correct, this would be a new explicit enumerator for a special two-type ASEP, connecting a concrete Markov-process statistic (switches of the 1 and 2 particles) to rational generating functions. The paper is commendably explicit and builds transparently on cited prior work, with no fitted parameters. However, the central decomposition is not proved, and the displayed identities in Theorems 3.2 and 3.3 are internally inconsistent as written. These issues prevent the manuscript from supporting its main claim in its present form.
major comments (4)
- [Section 4, Theorem 4.1] The central decomposition is asserted without proof. The only justification given for restricting 1/2 swaps to returns to (m,0) is the sentence "As this can only happen when the path returns to (m,0)." No bijection or path-by-path argument is provided, and the independence of the segments counted by G_{m,0,m,0}(1,1;t) and G_{m,0,0,m}(1,1;t) is not established. Thus the expression for Aλ is not shown to count the intended ASEP paths; if the decomposition is false, Theorem 4.1 counts a different object.
- [Section 3, Theorems 3.2 and 3.3] The left-hand sides of Theorems 3.2 and 3.3 are G_{m,0,m,0}(1,1;t) and G_{m,0,0,m}(1,1;t), which are functions of t only, but the right-hand sides explicitly contain the variables x and y. For instance, Theorem 3.2 states G_{m,0,m,0}(1,1;t) = (1 - x^{m+1} + y G_{m,0}(x,y;t))/G_{m+1,0}(x,y;t), and Theorem 3.3 has an analogous dependence on x and y through G_{0,m}(x,y;t) and G_{m+1,0}(x,y;t). Unless x=y=1 is intended on the right-hand side, these are not valid identities; the paper does not state or justify such an evaluation.
- [Section 4, paragraph before Theorem 4.1] The sentence "This assumes we only allow 1 and 2 to switch places in one direction" contradicts Definition 2.1, which assigns nonzero transition probabilities (t/n and 1/n) to swaps in both directions between adjacent unequal species. If the intended model forbids one direction of swap, the process is not the ASEP defined in Definition 2.1 and must be redefined; otherwise, the restriction is unjustified and the roles of d and 1/d as marking the two directions of the same ASEP transition are unclear.
- [Section 3, Figure 1 and Section 4] The paper uses the "simplified linear version of the simplex" as the lattice for all generating functions G_{u,v}, and these are subsequently combined with Elizalde's bijection to count ASEP paths. The paper does not justify that path counts in this simplified linear lattice agree with the counts of walks in the simplicial regions used by Elizalde. Section 5's remark that Mortimer and Prellberg offer only partial results even for a triangle indicates that the simplification is nontrivial; a proof or precise reference for the equivalence is needed for the main argument to be sound.
minor comments (6)
- [Abstract] The phrase "a new power series results" should be "a new power series result," and "the two type ASEP" would be better as "the two-type ASEP."
- [Introduction] The word "probabilites" is a typo for "probabilities."
- [Figure 1] The caption and the numbers "(0, 6)(6, 0)" and "10020000 / 01000002" are not explained; the reader cannot tell what the simplified lattice represents or how the labels encode configurations.
- [Theorems 3.2 and 3.3] Both theorems refer to "Figure 4," but the manuscript contains only Figure 1; the reference should be corrected.
- [Definition 2.1] The transition probabilities use a factor 1/n and t/n, but n is not defined; presumably n is the number of lattice sites or the length of λ, and this should be stated explicitly.
- [Theorem 4.1] The initial condition "starting with 1, 2 adjacent" is not tied to the standard ASEP(λ) starting word 2100...0, and the relationship between the lattice coordinate (m,0) and the ASEP configuration is not spelled out.
Circularity Check
No circularity: Theorem 4.1 is a synthesis of external explicit walk counts; Section 4's unproved decomposition is a correctness gap, not a circular reduction.
full rationale
The derivation chain is external and non-circular. Theorem 3.1 is an explicit formula from Mortimer and Prellberg [3] for walks in the simplified linear lattice; Theorems 3.2 and 3.3 are obtained by standard first-return decompositions of those walks and by rearranging or solving the resulting identities, not by assuming the target formula. Section 4 then builds the two-type ASEP generating function Aλ out of the one-type G-functions using a path grammar in which excursions between visits to (m,0) carry d or 1/d marks for 1/2 switches. No parameter is fitted to data, no quantity is defined in terms of the claimed output, and no load-bearing assertion is justified solely by a self-citation; all cited external results are by other authors. The sentence 'As this can only happen when the path returns to (m,0)' in Section 4 is an unproved combinatorial bridge and is a genuine correctness risk, as is the later admission 'This assumes we only allow 1 and 2 to switch places in one direction,' which appears to diverge from Definition 2.1. These are missing proofs or model mismatches, not reductions of the conclusion to its inputs. Consequently, the paper has no significant circularity; the appropriate score is 0.
Assumptions & free parameters
assumptions (4)
- domain assumption Mortimer-Prellberg explicit formula for walks in a triangular/linear lattice (Theorem 3.1)
- domain assumption Elizalde's bijection between ASEP paths and walks in simplices
- ad hoc to paper The simplified linear lattice is equivalent to the full simplex for the counts used
- standard math The Markov process definition of ASEP with transition probabilities as given
Cite this review
Pith. "Pith review of An explicit power series result for the two type ASEP." pith.science (2026). https://pith.science/paper/7A2RWYAL
@misc{pith2026250722296,
author = {Pith},
title = {Pith review of: An explicit power series result for the two type ASEP},
year = {2026},
howpublished = {\url{https://pith.science/paper/7A2RWYAL}},
note = {Machine review of arXiv:2507.22296}
}
read the original abstract
Research in combinatorics has often focused on the ASEP (asymmetric simple exclusion process). The ASEP is inspired by processes in statistical mechanics, and involves particles of various species moving around a lattice. The particles do not change species. In the present paper, based on earlier results of Mortimer and Prellberg and others, we obtain a new power series results for the two type ASEP.
Reference graph
Works this paper leans on
-
[2]
Walks in simplices, cylindric tableaux, and asymmetric exclusion processes
S. Elizalde. “Walks in simplices, cylindric tableaux, and asymmetric exclusion processes”. Séminaire Lotharingien de Combinatoire 86B (2022)
work page 2022
-
[3]
On the Number of Walks in a Triangular Domain
P . Mortimer and T. Prellberg. “On the number of walks in a triangular domain”. The Electronic Journal of Combinatorics 22 (2015). arXiv:1402.4448
work page Pith review arXiv 2015
-
[1]
From multiline queues to Macdonald poly- nomials via the exclusion process
S. Corteel, O. Mandelshtam, and L. Williams. “From multiline queues to Macdonald poly- nomials via the exclusion process”. Proceedings of the 31st Conference on Formal Power Series and Algebraic Combinatorics 82B (2019), pp. 1–12. arXiv:1811.01024
arXiv 2019
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.