$q$-Deformed Discrete Whittaker Processes

Richard Ninness

REVIEW 4 major objections 4 minor 12 references

$q$-Deformed Discrete Whittaker Processes

T0 review · 4 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read A q-deformed Markov process on reverse plane partitions has Markov projections to every skew shape.

desk verdict Genuine q-deformation of O'Connell's discrete Whittaker processes with the right architecture, but the load-bearing appendix algebra is partially deferred; worth serious refereeing. read the letter →

arxiv 2509.02881 v2 pith:FTFP65XL submitted 2025-09-02 math.PR

Richard Ninness This is my paper · ORCID

classification math.PR MSC 60J2505A3033D80

keywords q-deformedTodalatticediscreteWhittakerprocessesMarkovprojectionsreverseplanepartitionsskewshapesMarkov-Doobtransformquantumgroupsq-binomialcoefficients

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

This paper introduces a continuous-time Markov process on non-negative integer arrays, which become reverse plane partitions when the drift parameters vanish, with jump rates deformed by a parameter q. It shows that for every skew shape λ/μ with μ contained in λ°, there is such a process whose projection onto the boundary λ/μ is again Markovian, with an explicit generator. The construction is driven by an intertwining relation for the quantum difference Toda Hamiltonian, and its q→1 limit recovers the previously known undeformed discrete Whittaker processes after rescaling. The result matters because it transfers the Markovian-projection structure of discrete Whittaker processes into the q-world tied to Whittaker functions of U_q(sl_{r+1}).

What carries the argument

The argument runs through the quantum difference Toda Hamiltonian H_q^r, its eigenfunction coefficients a_r(n;q) built from q-binomial weights, and the intertwining relation h_r∘q_r = q_r∘h_{r-1} (Theorem 3.1). From this intertwining a Doob transform L_r = a_r^{-1}∘h_r∘a_r yields the Markov generator, and the projection property is carried by the Markov-function criterion. For general skew shapes, Lemma 5.6 supplies the algebraic closure identity (34), proved by decomposing the region between λ and μ into hook shapes and telescoping sums of rate differences.

What would settle it

Evaluate identity (34) numerically for the smallest non-staircase skew shape, e.g. λ=(2,2) and μ=(1), with q distinct from 1 and generic drift parameters: pick an array π, let σ be its boundary restriction, compute both sides directly, and check equality. Any mismatch would disprove Lemma 5.6 and hence Theorem 1.2.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.2: for each skew shape λ/μ with μ⊂λ°, there exists a Markov process π(t) on the q-constrained array space Π^{λ,α} whose generator is G^{λ,α}_μ and whose initial conditional law is proportional to W̃_{λ,μ}(π,z;q). Under this evolution, the boundary values π(t)|_{λ/μ} form a Markov process on Π^{λ/μ,α} with explicit generator L^{λ/μ,α}=A^{-1}_{λ,μ}∘H^{λ/μ,α}∘A_{λ,μ}. Unlike the undeformed model, the parent process depends on the chosen skew shape: different skew shapes may need different parent processes, although all of them converge to the same process in the q→1 limit.

Load-bearing premise

The load-bearing premise is the algebraic closure identity (34), which balances boundary rate differences plus a potential against interior adjoint rate differences; if this identity fails, the Markov projection theorem collapses.

Editorial extensions

If this is right

If Theorem 1.2 is correct, every skew shape admits a q-deformed Whittaker-type Markov process with explicit boundary generators.
The q→1 limit recovers the existing discrete Whittaker processes after the time rescaling t→(1−q)²t.
The boundary process has an explicit Doob-transformed generator obtained from a q-Toda potential, giving concrete transition-rate formulas.
The conditional law of the full array given the boundary history is governed by the q-measure W̃_{λ,μ}, a q-analogue of the undeformed reverse-plane-partition measure.
The construction yields a probabilistic representation of coefficients of quantum-difference Toda eigenfunctions in the fundamental-Whittaker family.

Reading between the lines

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

The per-skew-shape construction suggests that a single q-deformed parent process projecting Markovianly onto all skew shapes simultaneously may not exist; a unifying object, if any, would likely be a q-Whittaker eigenfunction system rather than one process.
The hook-decomposition identity (34) may be reusable as a combinatorial lemma for other q-deformed particle systems on Young-diagram shapes with boundary interactions.
One testable extension is whether the q-deformed processes converge, under the appropriate scaling of q and time, to the q-TASEP or other q-Whittaker-driven interacting particle systems.
Since the weights involve q-binomial coefficients, the construction may connect to Macdonald-process-type deformations of the Toda lattice at specialization parameters.

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

You should know: this paper constructs the q-deformed analogue of O'Connell's discrete Whittaker processes, with Markov projection theorems for skew shapes. The architecture is standard and mostly executed correctly, and the central claims are plausible. But the main theorem rests on an appendix that leaves the hardest algebra unverified — degenerate hook cases are explicitly 'left to the reader' and over-counted terms are dismissed by assertion. That is the real soft spot.

What is actually new: the generators with q^{π_{i+1,j-1}−π_{ij}} nearest-neighbor interactions that vanish as q→1, the weight W̃_{λ,μ} with its quadratic phase D_{λ,μ}, and the per-skew-shape projection results. The paper is honest about the main weakening compared to O'Connell: the parent process depends on μ, so there is no single process projecting to every skew shape. Appendix B provides a partial q→1 consistency check, which is genuinely useful.

What the paper does well: the structure is clear (staircase case, then drifts, then general μ), the proof strategy is the standard intertwining/Doob/Rogers-Pitman machine, and the connection to U_q(sl_{r+1}) Whittaker vectors is the right motivation.

Soft spots, in proportion: - Theorem 3.1's key identity is cited from the author's own [9], Lemma 5.1, without proof. A generalized version is proven later as Lemma 4.1, but the author never connects it to the cited lemma. Easy fix, but the foundational intertwining should not rest on an unshown citation. - Lemma 5.5.1, equation (33), drops the W̃ factor on the right side; the warm-up version (24) has it. Almost certainly a typo, but it indicates the algebra has not been checked line by line. - Appendix A: the proof of Lemma 5.6 works out nondegenerate hooks in detail, then says degenerate hooks are 'left to the reader to confirm' and, in the final paragraph, discards over-counted terms. These are exactly the places where a cancellation of q-exponentials could fail for non-staircase μ. I cannot exhibit a failure, and the structure is sound, but identity (34) is load-bearing and is not fully demonstrated. There is no machine-checked or code-based check of this algebra; it rests on hand bookkeeping. - Mechanical errors: references to a nonexistent Lemma 3.1 and Theorem 2.2. Rushed final pass.

Who it is for: specialists in integrable probability and Whittaker processes. It deserves a serious referee, with a request to fill in the degenerate-hook cases and connect Lemma 4.1 to the cited Lemma 5.1. If the Appendix A algebra closes, the central claim stands.

Referee Report

4 major / 4 minor

Summary. The paper introduces a q-deformed analogue of O'Connell's discrete Whittaker processes. The objects are continuous-time Markov processes on non-negative integer arrays (reverse plane partitions and their drift-perturbed analogues) on Young diagrams and skew shapes, with jump rates depending on q and on drift parameters. The central claim (Theorem 1.2, restated as Theorem 5.4) is that for every skew shape λ/μ with μ⊂λ°, one can choose a parent process on Π^{λ,α}, with generator G^{λ,α}_μ and initial conditional law proportional to W̃_{λ,μ}, such that the boundary values on λ/μ evolve as a Markov process with an explicit generator L^{λ/μ,α}. The proof strategy follows the classical route: an intertwining between quantum-difference Toda Hamiltonians, a Doob transform, and a Markov-functions criterion. The q→1 limit is claimed to recover O'Connell's processes. The main technical work is algebraic: proving two identities, Theorem 3.1 (an intertwining) and Lemma 5.6 (a telescoping identity in Appendix A), that make the Doob-transform argument close.

Significance. If the main theorem is correct, this is a meaningful contribution: it provides a one-parameter q-deformation of discrete Whittaker processes with explicit generators and weights, preserves Markov projections to skew shapes, and connects to quantum-group Whittaker functions and the q-difference Toda lattice. The paper also contains useful explicit formulas, a clear q→1 limit, and a coherent overall architecture based on intertwinings and Doob transforms. However, the significance is conditional on the central algebraic identities being valid. The manuscript currently contains several load-bearing gaps: one stated identity appears to be missing a factor, and the decisive lemma in Appendix A is completed only up to deferred degenerate cases and asserted cancellations. These are not cosmetic issues, because the main theorem is exactly the intertwining HΛ=ΛG. For these reasons the paper is not yet in publishable form, although the strategy is credible and the gaps may be fixable.

major comments (4)

[§5.2, Lemma 5.5.1 and Eq. (33)] The statement of Lemma 5.5.1 reads ̃b_{ij}(σ,z;q) L_{σ_{ij}} ̃W_μ(π,z;q)= ̃b_{ij}(π,z;q). As written this is dimensionally inconsistent: the left-hand side is a product of a rate and W̃, while the right-hand side has no W̃. In the proof of Theorem 5.4 the displayed computation replaces the sum of [L_{σ_{ij}} ̃W] ̃b_{ij}(σ) D_{σ_{ij}}F by ̃W ̃b_{ij}(π) D_{σ_{ij}}F, which requires W̃ on the right-hand side. The proof in the lemma only computes a difference involving D_{λ,μ} and does not track the q-Pochhammer and z-factor ratios in W̃. This is a load-bearing equality for the reduction H^{λ/μ,α}Λ=ΛG^{λ,α}_μ. Please correct the statement and supply the full computation, or explain explicitly the intended identity.
[Appendix A, Lemma 5.6 and Eq. (34)] The proof of Lemma 5.6 is the central algebraic step: identity (34) is exactly what makes the intertwining H^{λ/μ,α}Λ=ΛG^{λ,α}_μ close. The appendix reduces this to hook-by-hook telescoping, but the reduction is not complete. After Eq. (55), degenerate hooks are dismissed with 'we leave this to the reader to confirm', and the final paragraph states that 'one may take the sum ... and discard terms that were over-counted' and that 'intersections between staircases ... can be completely neglected'. These are precisely the places where cancellations could fail for non-staircase μ. The over-counting and intersection terms are asserted rather than shown to vanish. This is not a presentation issue: if (34) fails, the main theorem does not follow. Please provide a complete proof of Lemma 5.6, including all degenerate hook cases and an explicit accounting of every discarded or intersection term.
[§3, Theorem 3.1] The proof of Theorem 3.1 relies on the equality h^n_r q_r(n,k)=(h^{r-1}_k)^* q_r(n,k), cited from Lemma 5.1 of the author's paper [9]. The manuscript does not reproduce the lemma or prove the unparameterized case. Lemma 4.1 proves a generalized identity, but the reduction is not shown. Since this intertwining is foundational for the Doob transform and for all later results, the paper should be self-contained here: either state and prove the exact identity used in Theorem 3.1, or give a precise derivation from Lemma 4.1. A bare citation to an unpublished or concurrent preprint is not sufficient for a load-bearing step.
[§3, Eq. (13)] The decomposition ∑_{π∈Π^n_r} q^{∑ π_{ij}(π_{ij}-π_{i+1,j-1})} w_r(π;q)F(π)=q̃_r q̃_{r-1}⋯q̃_1 F_0 is asserted with 'leave it to the reader to confirm'. This identity underlies the proof of Theorem 3.2 and hence the Markov property of the outermost diagonal. The intended F_0 and the iterated summation structure are not defined. Please provide a construction of F_0 (or a clear bijection with the layers of reverse plane partitions) and a proof of the equality, since the subsequent intertwining computation depends on it.

minor comments (4)

[§5.2, Eq. (30)] The potential V_{λ,μ}(σ,z;q) is written with π variables in the second and fourth lines (q^{-π_{i-1,j+1}}, q^{π_{i,μ_i+1}}). Since V is supposed to be a function of the projected boundary values σ, this is at least a notational inconsistency and should be corrected.
[§3, proof of Theorem 3.2] The proof refers to 'Theorem 2.2' when invoking the Rogers–Pitman Markov-functions criterion; the intended reference appears to be [12] or Theorem 3.2 itself. Please fix the cross-reference.
[Appendix A, Eq. (54)] The left-hand side of Eq. (54) contains ̃b_{i-1,j+ℓ_{ij}+1}(π)−̃b_{i-2,j+ℓ_{ij}+1}(σ), which looks like an indexing typo. In a proof whose entire content is systematic bookkeeping, such typos make independent verification difficult; please proofread all long formulas in the appendix.
[§5.2, definitions before Eq. (29)] The seven auxiliary rate functions f^±, g^±, h^±, m are introduced abstractly, and only later are they specialized. It would help the reader to state immediately which specializations are used in the main theorem and to explain why the general notation is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the construction is a legitimate ansatz-based proof; the main theorem rests on algebraic identities whose proof is incomplete rather than circular.

full rationale

The paper does not fit parameters to data and then rename the fit a prediction; no empirical quantity is predicted. The rate functions (f+, f-, g+, g-, h+, h-, m) and the potential V are chosen in Section 5.2 precisely so that the closure identity (34) holds, but this is a standard reverse-engineered Doob h-transform construction, not a circular reduction: the theorem asserts that the constructed process has a Markov projection, and the proof verifies the intertwining HΛ = ΛG. The only self-citation is Lemma 5.1 of [9] used in Theorem 3.1; the paper proves a generalized version as Lemma 4.1, so the load-bearing identity is supported within the paper itself. The proof of Lemma 5.6 contains explicit gaps — degenerate hooks are 'left to the reader to confirm' (after (55)) and over-counted terms are 'discard[ed]' (final paragraph of Appendix A) — and Lemma 5.5.1 appears to omit a Wtilde factor on the right-hand side of (33). These are correctness risks and verification gaps, not circular reductions. Under the rule that circularity requires exhibiting a reduction by definition or a fitted input renamed as prediction, no such step is present.

Assumptions & free parameters 4 free parameters · 6 assumptions · 2 invented entities

The central claim rests on (i) the algebraic identities that make the intertwinings close (Lemma 5.1 of [9], re-proven in generalized form as Lemma 4.1 here); (ii) the coefficient formulas for a_r(n;q) from [9]; (iii) the Rogers-Pitman criterion; (iv) the implicit well-definedness of the processes. The seven specialized rate functions and the potential V_{lambda,mu} are ad hoc construction choices, not externally justified entities. No free parameters are fitted to data; q and the drifts are model inputs.

free parameters (4)

deformation parameter q
Model dial; fixed throughout; not fitted. The q to 1 limit recovers [11].
drift parameters alpha_1,...,alpha_r (z_k = q^{alpha_k})
Arbitrary nonnegative integers; inherited from [11]; the theorems hold uniformly in them.
specialized rate functions g+(pi)=pi_{i,j-1}-pi_{i-1,j+1}-beta_{i+1,j-1}, h+(pi)=-pi_{i-1,j+1}, m(pi)=pi_{i,j-1}-pi_{i+1
Chosen ad hoc in Section 5.2 so that closure identity (34) holds; these choices define the process being constructed.
potential V_{lambda,mu}(sigma,z;q)
Defined in (30) so that (G^{lambda/mu,alpha}+V_{lambda,mu})A_{lambda,mu}=0; reverse-engineered to make the Doob transform a generator.

assumptions (6)

domain assumption Lemma 5.1 of [9]: h^n_r q_r(n,k) = (h^{r-1}_k)^* q_r(n,k)
Invoked without proof in Theorem 3.1; self-cited from the author's [9]; generalized version Lemma 4.1 is proven in this paper.
domain assumption Coefficient recursion (8) and formula (7)/(18) for a_r(n;q)
Taken from [9]; 'We will make essential use of (8)'; no derivation in this paper.
standard math Rogers-Pitman Markov function criterion (Equation (1) in [12]) converts the intertwining into a Markov projection
Invoked in Theorem 3.2 and repeatedly afterwards to conclude the projection is Markovian.
domain assumption Existence of the continuous-time Markov process with the given generator and positive rates
The paper defines generators and asserts the processes; boundedness of rates and absorption at 0 are not proven, standard for such constructions.
standard math q-binomial / q-Pochhammer telescoping identities
Routine algebra used throughout Sections 3 to 5 and Appendix A.
domain assumption A_{lambda,mu}(sigma) > 0 so the conditional law K is well-defined
Weights are sums of nonnegative terms; positivity is implicit.

invented entities (2)

auxiliary rate functions f+, f-, g+, g-, h+, h-, m_{ij}(pi)
purpose: Define the generator G^{lambda,alpha}_mu of the process so that the intertwining H Lambda = Lambda G and identity (34) hold
Ad hoc to the paper; no external falsifiable handle; their specializations encode the model.
potential V_{lambda,mu}(sigma,z;q)
purpose: Added to the projected generator so that (G^{lambda/mu,alpha}+V)A_{lambda,mu}=0 and the projected Doob transform is a Markov generator
Chosen by hand in (30) to close the algebra; no independent justification outside the proof.

how reviews work

0 comments

Cite this review

Pith. "Pith review of $q$-Deformed Discrete Whittaker Processes." pith.science (2026). https://pith.science/paper/FTFP65XL

@misc{pith2026250902881,
  author       = {Pith},
  title        = {Pith review of: $q$-Deformed Discrete Whittaker Processes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FTFP65XL}},
  note         = {Machine review of arXiv:2509.02881}
}

abstract

We consider a Markov process on non-negative integer arrays of a certain shape, this shape being determined by general parameters in the model which correspond to drifts. In the case where these drifts are trivial, the arrays are reverse plane partitions. This model is closely related to the Whittaker functions of the quantum group $U_q(\mathfrak{sl}_{r+1})$ and the Toda lattice. Moreover, the Markov process we introduce can be viewed as a one-parameter deformation of the discrete Whittaker processes introduced by Neil O'Connell in [11]. We show that the $q$-deformed Markov process has non-trivial Markov projections to skew shapes.

Figures

Figures reproduced from arXiv: 2509.02881 by the authors.

**Figure 2.** Depicted is a staircase δr+1, red arrows for nearest-neighbor interactions between adjacent diagonals, blue arrow for nearest-neighbor interaction within an individual diagonal. The above can be generalized to a Markov process on nonnegative integer arrays (πij )(i,j)∈λ satisfying πij ≥ πi,j−1, πi−1,j − αi − ⋅ ⋅ ⋅ − αi+j−1, where λ is a fixed Young diagram. Let ℓ(λ) denote the length of the Young diagram λ, i.e. λi … view at source ↗

**Figure 3.** Depicted is a Young diagram λ with µ ⊂ λ ○ . The grey border signifies the boundary of λ/µ, the red shaded squares correspond to points in Vertµ(λ). Our model is a q-deformation of the one considered in [11], with many of the relevant quantities converging—after proper rescaling—to those in [11] as q → 1. The interactions between nearest-neighbors across rows (πij )i+j=s, (πij )i+j=s+1 are the natural q-deformation … view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: The different possible Gij shapes for (i, j) ∈ O1. For a hook F with origin (i, j), vertical length ℓ, and horizontal length ω we define S (F) = {(a, b) ∈ Ji, i + ωij − 1K × Jj, j + ℓ − 1K ∶ a + b ≤ i + j + ℓ − 1}. We will also find it useful to define the set C̃(µ (k)…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[9]

On a specialization of Toda eigenfunctions

Labelle, Antoine. ”On a specialization of Toda eigenfunctions.” arXiv preprint arXiv:2502.10655 (2025). Inventiones Mathematicae 151.1 (2003)

work page Pith review arXiv 2025
[1]

Gerasimov, S

A. Gerasimov, S. Kharchev, D. Lebedev and S. Oblezin. On a Gauss-Givental representation of quantum Toda chain wave equation. Int. Math. Res. Not. (2006) 1–23 26

work page 2006
[2]

”Macdonald processes.” Probability Theory and Related Fields 158.1 (2014): 225-400

Borodin, Alexei, and Ivan Corwin. ”Macdonald processes.” Probability Theory and Related Fields 158.1 (2014): 225-400

work page 2014
[3]

Ethier, Stewart N., and Thomas G. Kurtz. Markov processes: characterization and convergence. John Wiley & Sons, 2009

work page 2009
[4]

Whittaker functions on quantum groups and q-deformed Toda operators

Etingof, Pavel. ”Whittaker functions on quantum groups and q-deformed Toda operators.” arXiv preprint math/9901053 (1999)

work page Pith review arXiv 1999
[5]

”Fermionic formulas for eigenfunctions of the difference Toda Hamiltonian.” Letters in Mathematical Physics 88 (2009): 39-77

Feigin, Boris, et al. ”Fermionic formulas for eigenfunctions of the difference Toda Hamiltonian.” Letters in Mathematical Physics 88 (2009): 39-77

work page 2009
[6]

Feigin, Boris, Michio Jimbo, and Tetsuji Miwa. ”Gelfand–Zetlin basis, Whittaker Vectors and a Bosonic Formula for the sln+1 Principal Subspace.” Publications of the Research Institute for Mathematical Sciences 47.2 (2011): 535-551

work page 2011
[7]

”On q-Deformed-Whittaker Function I.” Communications in Mathematical Physics 294.1 (2010): 97-119

Gerasimov, Anton, Dimitri Lebedev, and Sergey Oblezin. ”On q-Deformed-Whittaker Function I.” Communications in Mathematical Physics 294.1 (2010): 97-119

work page 2010

Show all 12 references

[8]

”Quantum K-theory on flag manifolds, finite-difference Toda lattices and quantum groups.”

Givental, Alexander, and Yuan-Pin Lee. ”Quantum K-theory on flag manifolds, finite-difference Toda lattices and quantum groups.”
[10]

”Directed polymers and the quantum Toda lattice.” (2012): 437-458

O’Connell, Neil. ”Directed polymers and the quantum Toda lattice.” (2012): 437-458

2012
[11]

”Discrete Whittaker processes.” Probability and Mathematical Physics 4.4 (2023): 965- 1002

O’Connell, Neil. ”Discrete Whittaker processes.” Probability and Mathematical Physics 4.4 (2023): 965- 1002

2023
[12]

Chris G., and J

Rogers, L. Chris G., and J. W. Pitman. ”Markov functions.” The Annals of Probability (1981): 573-582. 27

1981

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [9]

On a specialization of Toda eigenfunctions

Labelle, Antoine. ”On a specialization of Toda eigenfunctions.” arXiv preprint arXiv:2502.10655 (2025). Inventiones Mathematicae 151.1 (2003)

work page Pith review arXiv 2025

[2] [1]

Gerasimov, S

A. Gerasimov, S. Kharchev, D. Lebedev and S. Oblezin. On a Gauss-Givental representation of quantum Toda chain wave equation. Int. Math. Res. Not. (2006) 1–23 26

work page 2006

[3] [2]

”Macdonald processes.” Probability Theory and Related Fields 158.1 (2014): 225-400

Borodin, Alexei, and Ivan Corwin. ”Macdonald processes.” Probability Theory and Related Fields 158.1 (2014): 225-400

work page 2014

[4] [3]

Ethier, Stewart N., and Thomas G. Kurtz. Markov processes: characterization and convergence. John Wiley & Sons, 2009

work page 2009

[5] [4]

Whittaker functions on quantum groups and q-deformed Toda operators

Etingof, Pavel. ”Whittaker functions on quantum groups and q-deformed Toda operators.” arXiv preprint math/9901053 (1999)

work page Pith review arXiv 1999

[6] [5]

”Fermionic formulas for eigenfunctions of the difference Toda Hamiltonian.” Letters in Mathematical Physics 88 (2009): 39-77

Feigin, Boris, et al. ”Fermionic formulas for eigenfunctions of the difference Toda Hamiltonian.” Letters in Mathematical Physics 88 (2009): 39-77

work page 2009

[7] [6]

Feigin, Boris, Michio Jimbo, and Tetsuji Miwa. ”Gelfand–Zetlin basis, Whittaker Vectors and a Bosonic Formula for the sln+1 Principal Subspace.” Publications of the Research Institute for Mathematical Sciences 47.2 (2011): 535-551

work page 2011

[8] [7]

”On q-Deformed-Whittaker Function I.” Communications in Mathematical Physics 294.1 (2010): 97-119

Gerasimov, Anton, Dimitri Lebedev, and Sergey Oblezin. ”On q-Deformed-Whittaker Function I.” Communications in Mathematical Physics 294.1 (2010): 97-119

work page 2010