pith. sign in

arxiv: 2606.19689 · v2 · pith:EUFX6FHPnew · submitted 2026-06-18 · 💻 cs.HC

Syndesmoscope: The Power of Invariant Plots Linked to Traditional Network Views

Matt Oddo , Indira Sowy , Stephen Kobourov , Tamara Munzner This is my paper

Pith reviewed 2026-06-26 16:21 UTC · model grok-4.3

classification 💻 cs.HC
keywords network visualizationinvariant plotsgraph explorationlinked viewsforce-directed layoutsspectral bisectiondensity decompositioninteractive visualization
0
0 comments X

The pith

Linking force-directed network views with invariant plots via leapfrogging and hopscotching reveals patterns inaccessible through any single view.

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

The paper presents Syndesmoscope, a system that places a standard force-directed network diagram next to three invariant plots based on dense-sparse gradient, geodesic eccentricity, and spectral bisection, along with a newly introduced kSnakes plot derived from density decomposition. These panes are connected by leapfrogging, which links highlights across the differing visual patterns, and hopscotching, which extends selections by following edges in the underlying topology. Demonstrations across usage scenarios on a corpus of 72 diverse networks show that switching between the views and applying the interactions surfaces structures that stay hidden when any one representation is used alone. A sympathetic reader would care because traditional layouts produce inconsistent pictures of the same network while invariant plots supply stable references that can be contrasted directly.

Core claim

Syndesmoscope demonstrates that juxtaposing a force-directed layout with invariant plots derived from dense-sparse gradients, geodesic eccentricity, spectral bisection, and kSnakes density decomposition, combined with leapfrogging and hopscotching interactions, allows users to discover network patterns that cannot be seen in any individual view alone, as shown through usage scenarios on a corpus of 72 diverse networks.

What carries the argument

Leapfrogging and hopscotching, the linked highlighting and hop-based traversal interactions that connect the force-directed pane to the invariant plot panes.

If this is right

  • The same network topology produces multiple distinct yet consistent visual patterns that can be compared directly.
  • Selections made in one pane can be propagated to others to trace connectivity-based structures.
  • Invariant plots supply stable reference frames that offset the variability of force-directed layouts.
  • Patterns become visible only when the system switches emphasis between density, distance, spectral cuts, and decomposition views.

Where Pith is reading between the lines

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

  • The same linking mechanism could be applied to other pairs of variable and invariant representations in visualization.
  • The method reduces dependence on selecting a single optimal layout algorithm for analysis.
  • Quantitative measures of insight gain could be defined by counting newly visible substructures after each interaction.

Load-bearing premise

The four chosen invariant plots together with the two interactions are sufficient to surface otherwise inaccessible patterns in networks.

What would settle it

A network, whether in the 72-network corpus or similar, in which applying leapfrogging and hopscotching produces no additional visible structures beyond those already apparent in the force-directed view alone.

Figures

Figures reproduced from arXiv: 2606.19689 by Indira Sowy, Matt Oddo, Stephen Kobourov, Tamara Munzner.

Figure 1
Figure 1. Figure 1: Syndesmoscope shows four interpretable geometric layouts computed from the same graph. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Node-link views of the 38 labeled graphs of 4 nodes (black), with node positions fixed at the corners of a unit square to encode labels. Above [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The kSnakes invariant visual pattern of the ’Les Misérables’ literary [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Interpretable axes in Syndesmoscope panes. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The Syndesmoscope system architecture is divided into two parts, backend and frontend. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: The ’Square Grid’ example dataset: 196 nodes and 364 edges, a regular lattice of 14 nodes per side. We select the densest shell nodes in [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The ’Western Power Grid’ infrastructure dataset: 4941 substations (nodes) and 6594 transmission lines (edges), exhibits a counterintuitive [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Two separate instances of Syndesmoscope loaded with geometric datasets, each instance with only 2 of the 4 panes visible. On the left, [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The ’Authors (Network Science)’ social network dataset: 379 authors (nodes) and 914 collaborations (edges). In the AdjacencyMatrix, [PITH_FULL_IMAGE:figures/full_fig_p008_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: The ’Les Misérables’ literary dataset: 77 characters (nodes) and 254 spoken interactions (edges). In the kSnakes pane, we brush-select [PITH_FULL_IMAGE:figures/full_fig_p011_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: The ’Western Power Grid’ infrastructure dataset: 4941 substations (nodes) and 6594 transmission lines (edges). We continue the [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: The ’Fruitfly Brain’ neurological dataset: 2956 neurons (nodes) and 116922 synapses (edges). This large graph shows the scalability limits [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: The ’Fullerene (C2160)’ geometric dataset: 2160 carbon atoms (nodes) and 3240 bonds (edges). Selection of the high eccentricity polylines [PITH_FULL_IMAGE:figures/full_fig_p012_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: The ’Tree Graph (Binomial)’ simple dataset: 128 nodes and 127 edges. In the HopCensus pane, the shortest polylines at the top, which [PITH_FULL_IMAGE:figures/full_fig_p013_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: The ’Sunflower Seeds’ geometric dataset: 987 seeds (nodes) and 2924 neighbor connections (edges). This image is the result of a complex [PITH_FULL_IMAGE:figures/full_fig_p013_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: The ’Fullerene (C2160)’ geometric dataset: 2160 carbon atoms (nodes) and 3240 bonds (edges). With the diamond orientation of the [PITH_FULL_IMAGE:figures/full_fig_p014_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: The ’Social Network (Facebook)’ social network dataset: 2160 people (nodes) and 3240 friendships (edges). With the square orientation of [PITH_FULL_IMAGE:figures/full_fig_p014_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: The ’Game of Thrones’ literary dataset: 107 book characters (nodes) and 352 spoken interactions (edges). We can find the highest degree [PITH_FULL_IMAGE:figures/full_fig_p015_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: The ’London Underground’ infrastructure dataset: 369 stations (nodes) and 430 connections (edges). We select a long chain in the [PITH_FULL_IMAGE:figures/full_fig_p015_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: The ’Dolphins’ ecological dataset: 62 dolphins (nodes) and 159 interactions (edges). A brush selection of the upper half in the ForceDirected [PITH_FULL_IMAGE:figures/full_fig_p016_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: The ’Human Genome’ biological dataset: 3751 genes (nodes) and 9034 genetic interactions (edges). We select a high-degree node in the [PITH_FULL_IMAGE:figures/full_fig_p016_21.png] view at source ↗
read the original abstract

Traditional network representations, such as node-link views and adjacency matrices, can show dramatically different visual patterns, depending on the underlying layout or seriation algorithm. In contrast, invariant plots consistently surface the same visual pattern for the same input topology; yet researchers have underexplored them and have not integrated them into visualization systems. We present Syndesmoscope, an interactive system for network exploration that juxtaposes multiple views of the same network. Panes show a familiar a force-directed view alongside three panes with interpretable geometric layouts based on graph-theoretic properties: dense-sparse gradient, geodesic eccentricity, and spectral bisection. As a secondary contribution, we introduce kSnakes, a new invariant plot based on density decomposition. Syndesmoscope supports two key interactions: leapfrogging, or linked highlighting between different and interpretable visual patterns; and hopscotching, or hop-based traversal that extends data selections through the underlying topology. Through usage scenarios across a corpus of 72 diverse networks, we demonstrate how these interactions reveal network patterns inaccessible through any single view alone. Live demo available at https://syndesmoscope.vercel.app/.

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.

Referee Report

2 major / 2 minor

Summary. The manuscript presents Syndesmoscope, an interactive system juxtaposing force-directed node-link views with three invariant plots (dense-sparse gradient, geodesic eccentricity, spectral bisection) plus a new kSnakes plot based on density decomposition. It defines leapfrogging (linked highlighting across views) and hopscotching (topology-extended selection traversal) interactions and claims that usage scenarios on a corpus of 72 networks demonstrate patterns inaccessible via any single view.

Significance. If the linked invariant plots and interactions can be shown to surface genuinely new structural insights, the work would usefully extend network visualization practice by integrating interpretable geometric layouts with conventional views. The live demo at https://syndesmoscope.vercel.app/ is a concrete strength that aids reproducibility and allows direct inspection of the described interactions.

major comments (2)
  1. [Usage Scenarios] Usage Scenarios section: the central claim that leapfrogging and hopscotching 'reveal network patterns inaccessible through any single view alone' rests entirely on author-curated narrative scenarios across 72 networks. No quantitative metric of insight gain, no controlled comparison against single-view baselines, and no user study are reported, leaving the inaccessibility assertion untested.
  2. [System Design] System Design / kSnakes subsection: the manuscript introduces kSnakes as a new invariant plot but provides no formal definition, invariance proof, or comparison showing how its density-decomposition layout differs from or improves upon the three other invariant plots already included.
minor comments (2)
  1. [Abstract] The abstract and introduction should explicitly state the selection criteria or diversity metrics used for the 72-network corpus so readers can assess representativeness.
  2. [Figures] Figure captions for the invariant plots would benefit from explicit axis labels or legends indicating the plotted graph-theoretic quantities.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on Syndesmoscope. The comments highlight opportunities to clarify our evaluation approach and the presentation of kSnakes. We respond to each major comment below.

read point-by-point responses

  1. Referee: [Usage Scenarios] Usage Scenarios section: the central claim that leapfrogging and hopscotching 'reveal network patterns inaccessible through any single view alone' rests entirely on author-curated narrative scenarios across 72 networks. No quantitative metric of insight gain, no controlled comparison against single-view baselines, and no user study are reported, leaving the inaccessibility assertion untested.

    Authors: We agree that the central claim rests on author-curated scenarios without quantitative metrics, controlled comparisons, or a user study, leaving the assertion of inaccessibility untested in a formal sense. In the network visualization literature, usage scenarios are a standard method for demonstrating the utility of new linked views and interactions on diverse data. To address the concern, we will revise the abstract and Usage Scenarios section to replace 'demonstrate' with 'illustrate' and add an explicit limitations paragraph noting the illustrative nature of the scenarios and the value of future controlled studies. This is a partial revision, as a full user study lies outside the current scope. revision: partial

  2. Referee: [System Design] System Design / kSnakes subsection: the manuscript introduces kSnakes as a new invariant plot but provides no formal definition, invariance proof, or comparison showing how its density-decomposition layout differs from or improves upon the three other invariant plots already included.

    Authors: The kSnakes subsection currently describes the plot via density decomposition but lacks an explicit algorithmic definition and direct comparisons. We will expand the subsection to include a formal step-by-step definition of the kSnakes construction, a brief argument for invariance derived from the decomposition process, and side-by-side comparisons that position kSnakes as providing a distinct hierarchical density view relative to the dense-sparse gradient, geodesic eccentricity, and spectral bisection plots. These additions will appear in the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity: system design paper with no derivation chain

full rationale

The paper introduces a visualization system (Syndesmoscope) and two interactions (leapfrogging, hopscotching) demonstrated via narrative usage scenarios on 72 networks. No equations, parameters, or mathematical derivations exist that could reduce to inputs by construction. The central claim of revealing inaccessible patterns rests on author-selected scenarios rather than any self-referential definition, fitted prediction, or self-citation load-bearing step. This matches the default expectation of no significant circularity for non-derivational contributions.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is a systems and visualization paper with no mathematical derivations or fitted parameters. The ledger is empty because the central claims rest on design choices and demonstration rather than axioms or free parameters.

pith-pipeline@v0.9.1-grok · 5741 in / 1146 out tokens · 24815 ms · 2026-06-26T16:21:33.317045+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

40 extracted references · 1 linked inside Pith

  1. [1]

    Abdelaal, N

    M. Abdelaal, N. D. Schiele, K. Angerbauer, K. Kurzhals, M. Sedlmair, and D. Weiskopf. Comparative evaluation of bipartite, node-link, and matrix- based network representations.IEEE Trans. Visualization & Computer Graphics (TVCG), 29(1):896–906, 2022. 1

    2022

  2. [2]

    Alvarez-Hamelin, L

    J. Alvarez-Hamelin, L. Dall’Asta, A. Barrat, and A. Vespignani. Large scale networks fingerprinting and visualization using the k-core decom- position.Advances in neural information processing systems, 18, 2005. 4

    2005

  3. [3]

    Auber, D

    D. Auber, D. Archambault, R. Bourqui, M. Delest, J. Dubois, A. Lambert et al. TULIP 5.Encyclopedia of social network analysis and mining, pp. 1–28, 2017. 4

    2017

  4. [4]

    J. P. Bagrow and E. M. Bollt. An information-theoretic, all-scales approach to comparing networks.Applied Network Science, 4(1):1–15, 2019. 2, 5, 9

    2019

  5. [5]

    J. P. Bagrow, E. M. Bollt, J. D. Skufca, and D. Ben-Avraham. Portraits of complex networks.Europhysics Letters, 81(6):68004, 2008. 2, 4, 5

    2008

  6. [6]

    Bastian, S

    M. Bastian, S. Heymann, and M. Jacomy. Gephi: An open source software for exploring and manipulating networks. InProceedings of the interna- tional AAAI conference on web and social media, vol. 3, pp. 361–362,

  7. [7]

    Batagelj, A

    V . Batagelj, A. Mrvar, et al. Pajek-program for large network analysis. Connections, 21(2):47–57, 1998. 4

    1998

  8. [8]

    Batagelj and M

    V . Batagelj and M. Zaversnik. An o(m) algorithm for cores decomposition of networks.arXiv preprint cs/0310049, 2003. 2, 3

    Pith/arXiv arXiv 2003

  9. [9]

    Behrisch, B

    M. Behrisch, B. Bach, N. Henry Riche, T. Schreck, and J.-D. Fekete. Matrix reordering methods for table and network visualization.Computer Graphics Forum (CGF), 35:693–716, 2016. 1, 3

    2016

  10. [10]

    S. P. Borgatti and M. G. Everett. Models of core/periphery structures. Social networks, 21(4):375–395, 2000. 4

    2000

  11. [11]

    Bostock, V

    M. Bostock, V . Ogievetsky, and J. Heer. D3: Data-driven documents.IEEE transactions on visualization and computer graphics, 17(12):2301–2309,

  12. [12]

    Chimani, C

    M. Chimani, C. Gutwenger, M. Jünger, G. W. Klau, K. Klein, P. Mutzel et al. The open graph drawing framework (OGDF).Handbook of graph drawing and visualization, 2011:543–569, 2013. 3

    2011

  13. [13]

    Concas, C

    A. Concas, C. Fenu, G. Rodriguez, and R. Vandebril. The seriation problem in the presence of a double Fiedler value.Numerical Algorithms, 92(1):407–435, 2023. 3

    2023

  14. [14]

    Cosmograph: Interactive network visualization

    Cosmograph. Cosmograph: Interactive network visualization. https: //cosmograph.app, 2022. 4

    2022

  15. [15]

    Csardi and T

    G. Csardi and T. Nepusz. The igraph software.Complex syst, 1695:1–9,

  16. [16]

    H. Cui, Z. Lu, P. Li, and C. Yang. On positional and structural node features for graph neural networks on non-attributed graphs.ACM CIKM, pp. 3898–3902, 2022. 2

    2022

  17. [17]

    Ghoniem, J

    M. Ghoniem, J. Fekete, and P. Castagliola. A comparison of the readability of graphs using node-link and matrix-based representations.IEEE Conf. Information Visualization (InfoVis), pp. 17–24, 2004. 2

    2004

  18. [18]

    Girvan and M

    M. Girvan and M. E. Newman. Community structure in social and bi- ological networks.Proceedings of the national academy of sciences, 99(12):7821–7826, 2002. 4

    2002

  19. [19]

    Hagberg, P

    A. Hagberg, P. J. Swart, and D. A. Schult. Exploring network structure, dynamics, and function using networkX. Technical report, Los Alamos National Laboratory (LANL), 2007. 3, 5

    2007

  20. [20]

    C. R. Harris, K. J. Millman, S. J. Van Der Walt, R. Gommers, P. Vir- tanen, D. Cournapeau et al. Array programming with NumPy.nature, 585(7825):357–362, 2020. 5

    2020

  21. [21]

    Hébert-Dufresne, J

    L. Hébert-Dufresne, J. A. Grochow, and A. Allard. Multi-scale structure and topological anomaly detection via a new network statistic: The onion decomposition.Nature Scientific Reports, 6(1):31708, 2016. 2, 3

    2016

  22. [22]

    Henry and J.-D

    N. Henry and J.-D. Fekete. MatrixExplorer: a dual-representation sys- tem to explore social networks.IEEE Trans. Visualization & Computer Graphics (TVCG), 12(5):677–684, 2006. 4

    2006

  23. [23]

    X. Hu, A. Lu, and X. Wu. Spectrum-based network visualization for topology analysis.IEEE computer graphics and applications, 33(1):58– 68, 2012. 3

    2012

  24. [24]

    J. D. Hunter. Matplotlib: A 2D graphics environment.Computing in science & engineering, 9(3):90–95, 2007. 3

    2007

  25. [25]

    Jamil, Y

    H. Jamil, Y . Liu, T. Caglar, C. Cole, N. Blanchard, C. Peterson et al. Ham- ming similarity and graph Laplacians for class partitioning and adversarial image detection. pp. 590–599, 2023. 3

    2023

  26. [26]

    Kairam, D

    S. Kairam, D. MacLean, M. Savva, and J. Heer. GraphPrism: Compact vi- sualization of network structure.Proceedings of the International Working Conference on Advanced Visual Interfaces, pp. 498–505, 2012. 4, 9

    2012

  27. [27]

    S. H. Lee, M. Cucuringu, and M. A. Porter. Density-based and transport-based core-periphery structures in networks.Physical Review E, 89(3):032810, 2014. 4

    2014

  28. [28]

    McNutt, M

    A. McNutt, M. C. Stone, and J. Heer. Mixing linters with GUIs: A color palette design probe.IEEE Trans. Visualization & Computer Graphics (TVCG), 31(1):327–337, 2024. 6

    2024

  29. [29]

    M. Newman. The structure and function of complex networks.SIAM Review, 45(2):167–256, 2003. 2

    2003

  30. [30]

    M. I. Oddo, S. Kobourov, and T. Munzner. The census-stub graph invariant descriptor.IEEE Trans. Visualization & Computer Graphics (TVCG), 31(3):1945–1961, 2025. 1, 2, 4, 6

    1945

  31. [31]

    M. I. Oddo, R. Smith, S. Kobourov, and T. Munzner. Visualization tasks for unlabeled graphs.IEEE Trans. Visualization & Computer Graphics (TVCG), 2026. 2, 4

    2026

  32. [32]

    G. A. Pavlopoulos, D. Paez-Espino, N. C. Kyrpides, and I. Iliopoulos. Em- pirical comparison of visualization tools for larger-scale network analysis. Advances in bioinformatics, 2017(1):1278932, 2017. 4

    2017

  33. [33]

    T. P. Peixoto. The Netzschleuder network catalogue and repository.Zen- odo, 2020. 6

    2020

  34. [34]

    Renoust, G

    B. Renoust, G. Melançon, and T. Munzner. Detangler: Visual analytics for multiplex networks. InComputer graphics forum, vol. 34, pp. 321–330. Wiley Online Library, 2015. 2, 4

    2015

  35. [35]

    J. C. Roberts. State of the art: Coordinated & multiple views in exploratory visualization. InFifth international conference on coordinated and mul- tiple views in exploratory visualization (CMV 2007), pp. 61–71. IEEE,

    2007

  36. [36]

    Shannon, A

    P. Shannon, A. Markiel, O. Ozier, N. S. Baliga, J. T. Wang, D. Ramage et al. Cytoscape: A software environment for integrated models of biomolecular interaction networks.Genome research, 13(11):2498–2504, 2003. 3

    2003

  37. [37]

    X. Shu, A. Pister, J. Tang, F. Chevalier, and B. Bach. Does this have a par- ticular meaning? Interactive pattern explanation for network visualizations. IEEE Trans. Visualization & Computer Graphics (TVCG), 31(1):677–687,

  38. [38]

    Wickham et al

    H. Wickham et al. Elegant graphics for data analysis, 2016. 3

    2016

  39. [39]

    Yoghourdjian, T

    V . Yoghourdjian, T. Dwyer, K. Klein, K. Marriott, and M. Wybrow. Graph Thumbnails: Identifying and comparing multiple graphs at a glance.IEEE Trans. Visualization & Computer Graphics (TVCG), 24(12):3081–3095,

  40. [40]

    edgiest edge

    yWorks GmbH. yWorks: Network & diagram visualization. https: //www.yworks.com, 2024. 3 A APPENDIX In this Appendix we present 12 more scenarios, in addition to the 6 usage scenarios in the main paper. Fig. 10: The ’Les Misérables’ literary dataset: 77 characters (nodes) and 254 spoken interactions (edges). In the kSnakes pane, we brush-select all the node...

    2024