arxiv: 2604.10497 · v2 · submitted 2026-04-12 · 💻 cs.ET · quant-ph

Recognition: unknown

Entangled happily ever after: Wedding reception seating mapped to classical and quantum optimizers

Karie A. Nicholas , Vikram Khipple Mulligan

Authors on Pith no claims yet

Pith reviewed 2026-05-10 16:34 UTC · model grok-4.3

classification 💻 cs.ET quant-ph

keywords wedding seating optimizationcost function networksquantum annealingD-Waveclassical Monte Carlooptimization benchmarksprotein design mappingsconstraint satisfaction

0 comments

The pith

Wedding seating optimization can be mapped to cost function networks, where classical Monte Carlo methods outperform D-Wave quantum annealing.

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

The paper shows that planning optimal seating for a wedding reception, with constraints on guest relationships, shared interests, and physical needs, can be expressed as a cost function network problem. This formulation permits direct testing of both classical solvers and quantum annealing techniques that use one-hot, domain-wall, and approximate binary mappings originally developed for protein design. In direct comparisons, classical Monte Carlo methods readily located optimal seating plans, while the D-Wave Advantage 2 system did not, prompting the authors to release the full benchmark set and conversion code as a Masala plugin for ongoing algorithm testing.

Core claim

The central discovery is that a real-world wedding reception seating problem was formulated as a cost function network using one-hot, domain-wall, and approximate binary mappings from protein design. Classical Monte Carlo optimization in the Masala suite found the globally optimal arrangements, whereas quantum annealing on the D-Wave Advantage 2 system failed to return those solutions despite succeeding on similarly structured protein design instances. The work supplies the benchmark problems and software plugin to support systematic comparisons of classical and quantum solvers on this class of practical constraint problems.

What carries the argument

Cost function network (CFN) representation of seating constraints, converted via one-hot, domain-wall, and approximate binary mappings to enable both classical Monte Carlo and quantum annealing solvers.

If this is right

This benchmark set can be used to evaluate future classical and quantum optimization algorithms on realistic social-constraint problems.
Performance gaps between classical Monte Carlo and quantum annealing may depend on the density and type of constraints present.
The released Masala plugin allows easy conversion of similar guest-arrangement tasks for systematic testing.
Quantum approaches may require additional refinement or tuning to reach the solution quality classical methods achieve on constraint-heavy planning tasks.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Optimization landscapes arising from interpersonal constraints may differ enough from molecular design problems to affect which solvers perform best.
Scaling the seating problem to larger events could help identify regimes where quantum annealing gains an advantage over classical methods.
The same mapping strategy could be applied to related planning tasks such as conference seating or logistics scheduling.

Load-bearing premise

The assumption that mappings developed for protein design problems transfer directly to wedding seating constraints without needing problem-specific tuning or that the chosen problem size and weights are representative of broader optimization performance.

What would settle it

Running the released seating benchmark instances on the D-Wave Advantage 2 and checking whether it returns the same optimal arrangements that classical Monte Carlo methods identify, or better ones, within comparable runtime.

Figures

Figures reproduced from arXiv: 2604.10497 by Karie A. Nicholas, Vikram Khipple Mulligan.

**Figure 1.** Figure 1: Function of the optimize_seating application in the Seating Optimization Masala Plugins library. This application permits a user to configure a seating optimization problem (salmon), converts it into a general CFN problem (brown), and encodes it as any compatible data representation available in a Masala plugin library (orange, green), permitting the optimization to be performed with any plugin Masala sol… view at source ↗

**Figure 3.** Figure 3: Seating solutions from classical Monte Carlo optimization. ( [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

read the original abstract

Although optimization is one of the most promising applications of quantum computers, the development of effective optimization strategies requires real-world test cases. When planning our recent wedding reception, we realized that the problem of optimally seating our guests, given constraints related to guests' relatedness, shared interests, and physical needs, could be mapped to a cost function network (CFN) form solvable with classical or quantum optimization algorithms. We compared the seating optimization performance of classical Monte Carlo CFN solvers in the Masala software suite to that of quantum annealing-based CFN optimization algorithms using one-hot, domain-wall, and approximate binary mappings, which we had developed for protein design problems. Surprisingly, the D-Wave Advantage 2 system, which performs well on similarly-structured CFN problems for protein design, struggled to return optimal seating arrangements that were easily found by classical Monte Carlo methods. We provide our seating optimization benchmark set, and code to convert seating optimization problems to CFN problems, as a plugin library for Masala, permitting this class of real-world problems to be used to benchmark performance of current and future classical CFN solvers, quantum optimization algorithms, and quantum computing hardware.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper's real contribution is a public seating benchmark and Masala plugin for CFN solvers, but the claim that D-Wave underperforms rests on unadjusted protein-design mappings that may not fit the new constraints well.

read the letter

The punchline is straightforward: this paper converts a wedding seating problem into a CFN benchmark, releases the instances and conversion code as a Masala plugin, and reports that classical Monte Carlo found good solutions while their D-Wave Advantage 2 runs did not. That benchmark release is the part worth paying attention to for anyone who tests optimization methods on real-world graphs with cardinality and pairwise terms. The authors reuse their earlier one-hot, domain-wall, and approximate-binary encodings from protein design work, which keeps the technical lift low but also limits how much is truly new in the algorithm sense. The concrete example and open data make it easy for others to add this problem class to their test suites without starting from scratch. That is useful even if the seating instance set stays small. The soft spot is the performance comparison. The abstract says D-Wave struggled on instances that Monte Carlo solved easily, yet the encodings and penalty weights were taken directly from protein design without apparent re-tuning for table-size constraints or the specific density and sign pattern of guest-pair terms. If those choices embed higher barriers or poorer minor embeddings on the hardware graph, the gap could be an artifact of the mapping rather than a general statement about quantum annealing on this problem type. Without numbers on run counts, success rates, or how the weights were set, it is hard to judge how representative the result is. The paper is aimed at researchers who benchmark classical and quantum CFN solvers on applied problems. A reader who needs fresh, non-synthetic instances with mixed hard and soft constraints will get immediate value from the released code and data. It deserves a serious referee because the benchmark itself is a legitimate applied contribution that stands on its own. Referees can push for clearer details on the mapping choices and quantitative results, but the core release is worth the time. I would send it for review rather than desk reject.

Referee Report

3 major / 2 minor

Summary. The manuscript maps a real-world wedding reception seating problem—with constraints on guest relatedness, shared interests, and physical needs—to a Cost Function Network (CFN) formulation. It compares classical Monte Carlo CFN solvers from the Masala software suite against quantum annealing on the D-Wave Advantage 2 system, using one-hot, domain-wall, and approximate-binary encodings previously developed for protein design. The central claim is that classical methods easily recover optimal seating arrangements while the quantum annealer struggles; the authors release the benchmark instances and a Masala plugin for converting seating problems to CFNs to support future benchmarking.

Significance. If the performance gap is substantiated, the work supplies a reproducible, real-world benchmark with heterogeneous cardinality and pairwise constraints that differs in structure from typical protein-design CFNs. The explicit release of the benchmark set and conversion code as open-source Masala plugin is a clear strength, enabling direct, falsifiable comparisons across classical solvers, quantum algorithms, and hardware. This contributes concrete test cases to the optimization literature rather than abstract claims.

major comments (3)

[Abstract and Results] Abstract and Results: The claim that 'D-Wave Advantage 2 ... struggled to return optimal seating arrangements that were easily found by classical Monte Carlo methods' is presented without any quantitative metrics (success probability, time-to-solution, number of samples, energy gaps, or statistical error bars). No tables or figures report these values for the different encodings or problem instances, rendering the magnitude and reliability of the performance difference unverifiable.
[Methods] Methods: The one-hot, domain-wall, and approximate-binary mappings (and their associated penalty weights) are imported directly from the authors' prior protein-design papers without re-optimization or sensitivity analysis for the seating problem's hard per-table cardinality constraints and heterogeneous pairwise terms. Because wedding seating introduces different density and sign patterns than contact potentials, the effective energy landscape and minor-embedding quality on the Pegasus/Zephyr topology may differ; this transfer assumption is load-bearing for attributing the observed gap to the annealer rather than to the encoding.
[Results/Discussion] Results/Discussion: No analysis is provided of chain-break statistics, embedding overhead, or the effect of the chosen constraint weights on the D-Wave samples. Without these diagnostics it is impossible to determine whether the reported difficulty arises from the problem class, the hardware topology, or the specific mapping parameters.

minor comments (2)

[Supplementary Material or Results] The manuscript would benefit from an explicit table listing the number of guests, table sizes, and the types/weights of constraints in each released benchmark instance.
[Methods] Notation for the CFN energy function could be clarified by writing out the penalty terms for the cardinality constraints alongside the pairwise terms.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their thoughtful and constructive comments, which have helped us improve the clarity and rigor of our manuscript. We address each major comment below and have made revisions to incorporate additional quantitative analyses and diagnostics.

read point-by-point responses

Referee: [Abstract and Results] The claim that 'D-Wave Advantage 2 ... struggled to return optimal seating arrangements that were easily found by classical Monte Carlo methods' is presented without any quantitative metrics (success probability, time-to-solution, number of samples, energy gaps, or statistical error bars). No tables or figures report these values for the different encodings or problem instances, rendering the magnitude and reliability of the performance difference unverifiable.

Authors: We agree that the original presentation of results lacked the quantitative metrics necessary to fully substantiate the claims. In the revised manuscript, we have added Table 2 reporting success probabilities, time-to-solution (TTS), number of samples, and energy gaps for each encoding across the benchmark instances, including statistical error bars from multiple runs. We have also included Figure 3 showing the distribution of solution energies for classical and quantum methods. These additions make the performance gap verifiable and quantify the difference. revision: yes
Referee: [Methods] The one-hot, domain-wall, and approximate-binary mappings (and their associated penalty weights) are imported directly from the authors' prior protein-design papers without re-optimization or sensitivity analysis for the seating problem's hard per-table cardinality constraints and heterogeneous pairwise terms. Because wedding seating introduces different density and sign patterns than contact potentials, the effective energy landscape and minor-embedding quality on the Pegasus/Zephyr topology may differ; this transfer assumption is load-bearing for attributing the observed gap to the annealer rather than to the encoding.

Authors: The referee is correct that the encodings were transferred from our prior work on protein design without specific re-optimization for the seating problem. While these mappings are intended to be general-purpose for CFN problems with cardinality constraints, we acknowledge the potential differences in constraint density. In the revision, we have added a sensitivity analysis on the penalty weights for the hard cardinality constraints and pairwise terms specific to the wedding seating instances. We report the results of this analysis in a new subsection of the Methods, showing that the chosen weights are robust within a reasonable range, and discuss any adjustments made. revision: yes
Referee: [Results/Discussion] No analysis is provided of chain-break statistics, embedding overhead, or the effect of the chosen constraint weights on the D-Wave samples. Without these diagnostics it is impossible to determine whether the reported difficulty arises from the problem class, the hardware topology, or the specific mapping parameters.

Authors: We concur that diagnostic information on the quantum annealing runs is essential for interpreting the results. We have revised the manuscript to include an analysis of chain-break statistics, reporting the average chain-break fraction and its correlation with solution quality for each encoding. We also detail the embedding overhead (number of physical qubits per logical variable) and provide a discussion on how the constraint weights influence the sampled energies. This new content is added to the Results section and helps attribute the performance differences more precisely. revision: yes

Circularity Check

0 steps flagged

Minor self-citation of protein-design encodings without load-bearing reduction of the central empirical comparison

full rationale

The paper cites its own prior work for the one-hot, domain-wall, and approximate-binary mappings (abstract: 'using one-hot, domain-wall, and approximate binary mappings, which we had developed for protein design problems'). This is a self-citation but does not make the central claim circular: the seating problem, its specific cardinality and pairwise constraints, the benchmark instances, and the observed performance gap versus classical Monte Carlo are independent inputs. No equation reduces to a fitted parameter defined by the prior papers, no uniqueness theorem is imported to force the result, and the comparison itself is externally falsifiable on the new benchmark set. The derivation chain therefore remains self-contained against external benchmarks, warranting only the minimal score for a non-load-bearing self-citation.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The approach rests on standard cost-function-network formulations and previously published mappings; no new free parameters, axioms, or invented entities are introduced in the abstract.

pith-pipeline@v0.9.0 · 5506 in / 1082 out tokens · 51387 ms · 2026-05-10T16:34:12.600988+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

16 extracted references · 4 canonical work pages · 1 internal anchor

[1]

Computational Design of Peptide-Based Binders to Therapeutic Targets

Vikram Khipple Mulligan and Parisa Hosseinzadeh. Computational Design of Peptide-Based Binders to Therapeutic Targets. In Swapnil V . Ghodge, Kaustav Biswas, and Andrei A. Golosov, editors,Approaching the Next Inflection in Peptide Therapeutics: Attaining Cell Permeability and Oral Bioavailability, volume 1417 ofACS Symposium Series, pages 55–102. America...

2022
[2]

Pierce and Erik Winfree

Niles A. Pierce and Erik Winfree. Protein design is NP-hard.Protein Engineering, 15(10):779–782, October 2002

2002
[3]

A. R. Leach and A. P. Lemon. Exploring the conformational space of protein side chains using dead-end elimination and the A* algorithm. Proteins, 33(2):227–239, November 1998

1998
[4]

A new framework for computational protein design through cost function network optimiza- tion.Bioinformatics (Oxford, England), 29(17):2129–2136, September 2013

Seydou Traor ´e, David Allouche, Isabelle Andr´e, Simon de Givry, George Katsirelos, Thomas Schiex, and Sophie Barbe. A new framework for computational protein design through cost function network optimiza- tion.Bioinformatics (Oxford, England), 29(17):2129–2136, September 2013

2013
[5]

Brian Kuhlman and David Baker. Native protein sequences are close to optimal for their structures.Proceedings of the National Academy of Sciences of the United States of America, 97(19):10383–10388, September 2000

2000
[6]

Tristan Zaborniak, Noora Azadvari, Qiyao Zhu, S. M. Bargeen A. Turzo, Parisa Hosseinzadeh, P. Douglas Renfrew, and Vikram Khipple Mulligan. The open-source Masala software suite: Facilitating rapid methods development for synthetic heteropolymer design.Methods in Enzymology, 723:299–426, 2025

2025
[7]

Weitzner, Andrew M

Vikram Khipple Mulligan, Hans Melo, Haley Irene Merritt, Stewart Slocum, Brian D. Weitzner, Andrew M. Watkins, P. Douglas Renfrew, Craig Pelissier, Paramjit S. Arora, and Richard Bonneau. Design- ing Peptides on a Quantum Computer, September 2019.bioRxiv, doi:10.1101/752485

work page doi:10.1101/752485 2019
[8]

Domain wall encoding of discrete variables for quantum annealing and QAOA.Quantum Science and Technology, 4(4):045004, August 2019

Nicholas Chancellor. Domain wall encoding of discrete variables for quantum annealing and QAOA.Quantum Science and Technology, 4(4):045004, August 2019

2019
[9]

Tristan Zaborniak.Toward quantum computational biomolecular struc- ture prediction. Ph.D. dissertation, University of Victoria, Victoria, British Columbia, Canada, 2025

2025
[10]

Lewis, Oliver F

Andrew Leaver-Fay, Michael Tyka, Steven M. Lewis, Oliver F. Lange, James Thompson, Ron Jacak, Kristian W. Kaufman, P. Douglas Renfrew, Colin A. Smith, Will Sheffler, Ian W. Davis, Seth Cooper, Adrien Treuille, Daniel J. Mandell, Florian Richter, Yih-En Andrew Ban, Sarel J. Fleishman, Jacob E. Corn, David E. Kim, Sergey Lyskov, Monica Berrondo, Stuart Ment...

2011
[11]

Weitzner, Steven M

Julia Koehler Leman, Brian D. Weitzner, Steven M. Lewis, Jared Adolf-Bryfogle, Nawsad Alam, Rebecca F. Alford, Melanie Apra- hamian, David Baker, Kyle A. Barlow, Patrick Barth, Benjamin Basanta, Brian J. Bender, Kristin Blacklock, Jaume Bonet, Scott E. Boyken, Phil Bradley, Chris Bystroff, Patrick Conway, Seth Cooper, Bruno E. Correia, Brian Coventry, Rhi...

2020
[12]

G., and Roy, A

Jun Cai, William G. Macready, and Aidan Roy. A practical heuristic for finding graph minors, June 2014. arXiv:1406.2741 [quant-ph]

work page arXiv 2014
[13]

Inhomogeneous driving in quantum annealers can result in orders-of-magnitude improvements in performance.Quantum Science and Technology, 5(3):035011, June 2020

Juan I Adame and Peter L McMahon. Inhomogeneous driving in quantum annealers can result in orders-of-magnitude improvements in performance.Quantum Science and Technology, 5(3):035011, June 2020

2020
[14]

Mulligan

Mohit Pandey, Tristan Zaborniak, Hans Melo, Alexey Galda, and Vikram K. Mulligan. Multibody molecular docking on a quantum annealer, October 2022. arXiv:2210.11401 [q-bio]

work page arXiv 2022
[15]

A Quantum Approximate Optimization Algorithm

Edward Farhi, Jeffrey Goldstone, and Sam Gutmann. A Quantum Approximate Optimization Algorithm.arXiv:1411.4028 [quant-ph], November 2014. arXiv: 1411.4028

work page internal anchor Pith review Pith/arXiv arXiv 2014
[16]

Mario Motta, Chong Sun, Adrian T. K. Tan, Matthew J. O’Rourke, Erika Ye, Austin J. Minnich, Fernando G. S. L. Brand ˜ao, and Garnet Kin- Lic Chan. Determining eigenstates and thermal states on a quantum computer using quantum imaginary time evolution.Nature Physics, 16(2):205–210, February 2020

2020