Pith. sign in

REVIEW 3 major objections 4 minor 37 references

Curvature heterogeneity shifts the disorder–order transition of polar active agents to stronger alignment and drives swarms to an equatorial belt.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 15:36 UTC pith:3DNMDI7N

load-bearing objection A clean geometric reformulation with a plausible but not yet airtight central claim: the equatorial localization is old news, the transition shift is new but rests on single-run data and unvalidated geodesic expansions. the 3 major comments →

arxiv 2512.16267 v1 pith:3DNMDI7N submitted 2025-12-18 cond-mat.soft

A geometric framework for curvature-dependent collective behavior of polar active agents on curved surfaces

Tatsuo Shibata This is my paper
classification cond-mat.soft
keywords active agentspolar alignmentcurved surfacesspheroidsdisorder-order transitionswarmingcollective motionparallel transport
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

Self-propelled agents on a curved surface must decide how to align their motion, and this paper shows that the surface's shape changes both the difficulty and the geometry of that decision. Using a minimal model of polar agents with a standard alignment rule, it finds a disordered-to-ordered transition on spheres and on oblate and prolate spheroids. The paper's main result is that as the surface departs from sphericity, the alignment strength needed to reach a given level of order increases, and strongly eccentric spheroids host swarms in an equatorial belt rather than near the poles. This matters because curved geometries are ubiquitous in biology—from cell surfaces to intestinal villi—and the framework provides a general way to see how curvature steers collective motion.

Core claim

The paper establishes that on spheroids of fixed surface area, increasing eccentricity e shifts the disorder–order transition for polar alignment to larger values of K: the K-value at a fixed polar order parameter rises by roughly 10% at e≈0.9 relative to a sphere. It further shows that for e≳0.5 the unit polar order vector P̂ is biased toward ±Z, so the swarm's plane of rotation becomes the equatorial plane and the poles are avoided. The temporal variance of P̂ is non-monotonic in e—first rising, then falling—which the paper interprets as a loss and then a recovery of orientational stability as curvature heterogeneity grows. The model treats curvature intrinsically through parallel transpor

What carries the argument

The framework expresses agent motion directly on the surface using stereographic coordinates and a metric-dependent equation of motion with active speed control, noise, and polar alignment. The alignment interaction requires comparing velocity vectors in different tangent planes, so the model parallel-transports each neighbor's velocity along the shortest geodesic to the focal agent; on spheres this transport is exact, while on spheroids it uses midpoint-based truncations of the geodesic distance and the transport map. The order parameter P is the area-weighted average of the normal to each agent's great-circle motion, giving both the degree and the orientation of polar order. These componen

Load-bearing premise

The central results rely on truncated midpoint formulas for geodesic distance and parallel transport, which assume the interaction radius is much smaller than the local curvature radius; for the highest eccentricities the two differ by only a factor of two to three in the very regions where the predicted swarms live.

What would settle it

Compute the same spheroid simulations using exact geodesic distances and parallel transport (obtained by numerically solving the geodesic boundary-value problem) for e=0.9, and compare the K-dependence of ∥P∥ and the distribution of P̂Z; if the upward shift of the transition and the equatorial bias vanish, the paper's central claim is refuted.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • On any surface with nonuniform Gaussian curvature, inducing a given level of polar order requires a stronger alignment interaction than on an equal-area sphere.
  • For spheroids with eccentricity above roughly 0.5, swarms form preferentially in the equatorial region, so their plane of rotation is the equator and the poles are depleted of swimmers.
  • The orientation of the polar order vector becomes biased along the symmetry axis for strongly eccentric spheroids, meaning curvature heterogeneity selects a preferred swarm orientation.
  • Because the method only needs a metric, it can be applied to arbitrary curved surfaces, not just spheres and spheroids, to predict where swarms will form.

Where Pith is reading between the lines

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

  • This suggests a geometric selection rule: polar swarms on a curved surface will be attracted to closed paths where the convergence of neighboring geodesics is weakest, which on spheroids is the equator; the same logic would predict specific swarm trajectories on tori or other surfaces with multiple candidate loops.
  • The non-monotonic variance of P̂ implies a crossover between a weak-coupling regime where curvature merely destabilizes the swarm direction and a strong-coupling regime where it acts as a pinning potential; the eccentricity at the crossover should shift if the interaction radius changes, a prediction the paper does not make.
  • Coupling this model to surface deformation could predict whether the equatorial-belt tendency of swarms reshapes the surface itself, connecting the ordering transition to tissue or vesicle morphogenesis.
  • The equatorial localization found here mirrors an effect reported for nematic active agents on spheroids, suggesting a common geometric origin for polar and nematic order on curved surfaces that could be tested with a single active-agent experiment on an ellipsoidal droplet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper introduces a geometric framework for active agents on curved surfaces, with a Vicsek-like polar alignment interaction implemented via parallel transport. After validating the single-agent dynamics on spheres and spheroids (uniform stationary distributions and a flat-space ABP-like MSD envelope), the author studies N=1000 interacting agents on spheres and on fixed-area oblate/prolate spheroids. The central claim is that, as eccentricity e increases, the disorder–order transition shifts to larger alignment strengths K and the swarming direction becomes biased toward the equatorial plane (Fig. 4e–h). A secondary claim is a non-monotonic spherical variance of the polar-order direction.

Significance. If correct, the paper provides a minimal, geometry-based simulation framework for polar active matter on curved surfaces and offers concrete, falsifiable predictions about curvature heterogeneity suppressing global order and orienting swarms. The single-agent sector is validated cleanly against analytic measures and the standard flat-space active-Brownian MSD, which gives confidence in the underlying stochastic-geometric integrator. The interacting results are emergent and not fitted, and the comparison across oblate and prolate spheroids is a useful step beyond spherical geometries. The main weakness is that the high-eccentricity conclusions rely on truncated geodesic/parallel-transport expansions that have not been benchmarked in the parameter regime where the effects are strongest.

major comments (3)
  1. [III.B, Eqs. (6)–(7) and Appendix D] The central quantitative conclusions—the upward shift of the transition point (Fig. 4f) and the equatorial-belt orientation (Fig. 4g)—are computed with midpoint expansions for the geodesic distance and parallel transport whose truncation errors are O(||Δ||^4) and O(||Δ||^3), respectively. For the e=0.9, R=10 spheroids, the smallest principal curvature radius is approximately 2.4 at the oblate equator and 3.1 at the prolate poles, only 2–3 times the interaction radius r=1. In the very regions where the reported effects occur, the discarded terms need not be small, and neighbor lists can be systematically distorted. Since the observed K-shift in Fig. 4f is about 10% at fixed order parameter, a systematic geometric bias of a few degrees in alignment could produce or inflate the effect. The paper should include a convergence test against exact spheroid geodesics and parallel transport (e.g.,
  2. [Fig. 4(e)–(f)] The reported shift of the transition point is extracted by interpolation of ∥P∥ curves at ∥P∥=0.2, 0.3, and 0.4, but no error bars or ensemble variability are given. For N=1000 and σ=0.2, finite-size fluctuations in ∥P∥ are substantial, and the differences between adjacent eccentricity curves in Fig. 4e appear small. Without a statistical measure—for instance, standard errors over independent initial conditions, or a bootstrap over runs—the monotonic increase of K with e is not established at the claimed precision. The authors should report error bars on the interpolated K values and show at least one check at different N or σ to confirm that the shift is not a finite-size artifact.
  3. [Fig. 4(g)–(h)] The abstract's claim that swarming 'localizes to an equatorial belt away from the poles' is supported only indirectly by the distribution of \hat P_Z, the component of the orbital-axis vector. While \hat P_Z near ±1 implies that the swarm's orbital plane is the equator, it does not directly quantify the spatial concentration of agents away from the poles. A direct measure—for example, steady-state agent density as a function of latitude, or the mean |Z| of agents in the ordered state—would make the localization claim explicit and more convincing. This is an easy addition that would substantially strengthen the paper.
minor comments (4)
  1. [III.B, text after Fig. 4(g)] Two figure cross-references appear to be wrong: the uniform distribution of \hat P_Z for the sphere is described as 'shown in Fig. 2(f)', and the biased \hat P_Z distributions are referenced as 'Fig. 4(f)'; both should be Fig. 4(g).
  2. [Fig. 4 caption] The caption contains the garbled text 'PCMBUF QSPMBUF', which appears to be leftover plotting labels. Please remove or replace it.
  3. [Fig. 4(e)–(h)] The simulation protocol for the interacting case should be stated more completely: how long are the runs, how is the transient removed, and are the 100 samples independent initial conditions or independent noise realizations? This is needed to assess the quoted spherical-variances and to reproduce the results.
  4. [Eq. (4)] The text states that the set N_i includes agent i itself, so for an isolated agent Eq. (4) contains a self-alignment term K v_i. Please state explicitly that this is intended and discuss its effect, especially in low-density regions where N_i is small. This is not necessarily a flaw, but it should be made explicit.

Circularity Check

0 steps flagged

No significant circularity: the transition and localization results are emergent simulation outputs, validated against external analytic benchmarks, with no fitted parameter renamed as a prediction and no load-bearing self-citation.

full rationale

The central results—the disorder–order transition on spheres and spheroids and the eccentricity-dependent shift and equatorial localization—are emergent outputs of the agent-based dynamics defined by Eqs. (1) and (4). Model parameters (K, σ, e) are scanned controls; the polar order parameter ∥P∥ is measured, not fitted, and no reported quantity is used to calibrate the model. The single-agent validation in Sec. II is checked against external analytic results: the flat-space active-Brownian MSD formula (Fig. 1c) and the exact uniform position distributions on sphere and spheroids, Eq. (3). The only self-citation, ref [34], supports the active-force form together with non-overlapping external references [32,33,35] and does not carry a load-bearing claim. The approximate geodesic-distance and parallel-transport expansions, Eqs. (6)–(7), are taken from the independent reference [37] and are explicitly labeled as approximations in Appendix D; their accuracy at large eccentricity is a numerical-error or correctness question, not a circularity, because the reported transition points and order parameters are not constructed to reproduce those expansions. The equatorial-swarming observation is openly acknowledged as previously reported in [22] and [8], so it is not presented as a novel derivation from this paper's own framework. No circular step in the derivation chain is identifiable.

Axiom & Free-Parameter Ledger

3 free parameters · 5 axioms · 0 invented entities

The central claim rests on (a) textbook surface geometry, (b) a borrowed active-motility form [32–35], (c) the modeling choice of parallel-transport alignment, and (d) an unvalidated numerical approximation at high eccentricity. No parameter is fitted to data; K, σ, and e are scanned control parameters. I=10, r/R=0.1, and the ‖P‖-level interpolation are hand choices that shape the reported transition thresholds.

free parameters (3)
  • Speed-maintenance coefficient I = 10 (dimensionless)
    Hand-set strength of the active speed-restoring term in Eqs. (1),(4); fixed at I=10 for all runs (Sec. III). Not scanned, and no sensitivity analysis is given, so the absolute transition thresholds K depend on an unexamined choice.
  • Interaction-to-system size ratio r/R = 0.1 (r=1 in rescaled units; R=10)
    Chosen via the space/time rescaling (Sec. III); determines mean neighbor number (~2.5 with N=1000) and the regime of validity of the midpoint geodesic approximations. Never varied.
  • Operational definition of the transition point = K at ||P|| = 0.2/0.3/0.4 via linear interpolation (Fig. 4f)
    The claimed shift of the transition point is read off interpolated levels of the order parameter rather than a thermodynamic or finite-size-scaling criterion; the shift magnitude (~10% in K) is a property of this definition.
axioms (5)
  • standard math Standard differential geometry of surfaces: the decomposition ∂²r/∂xᵢ∂xⱼ = Γ^k_ij ∂r/∂x_k + n h_ij (Eq. A5) yields the geodesic equation with forcing (Eq. 1).
    Appendix A: the derivation is textbook differential geometry; the paper's equation of motion is the covariant acceleration set equal to tangential forces.
  • standard math The single-agent stationary distribution is the uniform (Riemannian volume) measure on the sphere/spheroid, used to validate the sampler (Eq. 3, Fig. 2).
    Standard property of isotropic diffusion (plus speed maintenance) on a Riemannian manifold; the match with the analytic ρ(Z) is the empirical check.
  • domain assumption Self-propulsion is modeled by the friction-like active force I v(v₀² − |v|²), taken from active-Brownian and cell-migration models [32–35].
    Eq. (A10). Different motility mechanisms (fixed speed constraint, curvature-dependent speed, cuprotaxis) would change the clustering and transition; this one is assumed without comparison.
  • domain assumption Comparing velocities across tangent planes by parallel transport along the minimizing geodesic is the correct/adequate alignment rule.
    Sec. III, Eq. (4). The rule is motivated as 'fair, coordinate-independent', but no alternative (e.g., embed-and-align, chart-based alignment) is tested; the equatorial localization and transition shift could depend on this choice.
  • domain assumption The truncated midpoint expansions for distance and parallel transport (Eqs. (6)–(7), Appendix D, after Brewin [37]) are accurate enough at e up to 0.9.
    Where the local curvature radius (≈2.4 for oblate e=0.9 equator; ≈3.1 at prolate e=0.9 poles) is only 2–3× the interaction radius r=1, the O(‖Δ‖³/‖Δ‖⁴) truncation is unchecked; no convergence test against exact geodesics is reported.

pith-pipeline@v1.3.0-alltime-deepseek · 12873 in / 27469 out tokens · 257811 ms · 2026-08-03T15:36:55.293664+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of A geometric framework for curvature-dependent collective behavior of polar active agents on curved surfaces." pith.science (2026). https://pith.science/paper/3DNMDI7N

@misc{pith2026251216267,
  author       = {Pith},
  title        = {Pith review of: A geometric framework for curvature-dependent collective behavior of polar active agents on curved surfaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3DNMDI7N}},
  note         = {Machine review of arXiv:2512.16267}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In biological systems, active agents such as actomyosin and cells move and interact on curved surfaces, exhibiting diverse phenomena. These observations have motivated studies of how curvature shapes their collective behavior. Here, using a geometric framework, a minimal model is presented for interacting active agents on curved surfaces with Vicsek-like polar alignment. A transition between disordered and ordered states occurs on spheres as well as on oblate and prolate spheroids. As the deviation from sphericity increases, the transition point shifts to higher alignment strengths, and swarming localizes to an equatorial belt away from the poles, indicating that curvature heterogeneity influences the emergence of the polar-ordered state.

Figures

Figures reproduced from arXiv: 2512.16267 by Tatsuo Shibata.

Figure 1
Figure 1. Figure 1: FIG. 1. (a) A trajectory of an active agent on a sphere of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. (a,b) Active agent on oblate and prolate spheroids of [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. (a,b,c) Snapshot images of polar interacting active agents on sphere of radius [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. (a,b,c,d) Snapshot images of polar interacting active agents on oblate (a,b) and prolate (c,d) spheroids for three different [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

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

37 extracted references · 2 canonical work pages

  1. [1]

    D. S. Johnston and B. Sanson, Current Opinion in Cell Biology23, 540 (2011)

    2011

  2. [2]

    Schambergeret al., Advanced Materials35, e2206110 (2023)

    B. Schambergeret al., Advanced Materials35, e2206110 (2023)

    2023

  3. [3]

    Vikran and T

    E. Vikran and T. Hirashima, Seminars in Cell & Devel- opmental Biology173, 103633 (2025)

    2025

  4. [4]

    Arnold, I

    C. Arnold, I. Tahmaz, M.-L. Chapon, H. Maayouf, V. Luchnikov, J.-L. Milan, F. Borghi, and L. Pieuchot, BMC Biology23, 296 (2025)

    2025

  5. [5]

    S. C. Al-Izzi and R. G. Morris, Seminars in Cell & De- velopmental Biology120, 44 (2021)

    2021

  6. [6]

    F. C. Keber, E. Loiseau, T. Sanchez, S. J. DeCamp, L. Giomi, M. J. Bowick, M. C. Marchetti, Z. Dogic, and A. R. Bausch, Science345, 1135 (2014), 1409.1836

    Pith/arXiv arXiv 2014

  7. [7]

    C.-P. Hsu, A. Sciortino, Y. A. d. l. Trobe, and A. R. Bausch, Nature Communications13, 2579 (2022)

    2022

  8. [8]

    Clairand, A

    M. Clairand, A. Mozaffari, J. Hardo¨ uin, R. Zhang, C. Dor´ e, J. Ign´ es-Mullol, F. Sagu´ es, J. J. d. Pablo, and T. Lopez-Leon, Physical Review X14, 031049 (2024)

    2024

  9. [9]

    Nishiyama, T

    C. Nishiyama, T. Uesaka, T. Manabe, Y. Yonekura, T. Nagasawa, D. F. Newgreen, H. M. Young, and H. Enomoto, Nature Neuroscience15, 1211 (2012)

    2012

  10. [10]

    P´ erez-Gonz´ alez, G

    C. P´ erez-Gonz´ alez, G. Ceada, M. Matejˇ ci´ c, and X. Trepat, Current Opinion in Genetics & Development 72, 82 (2022)

    2022

  11. [11]

    W. Xi, S. Sonam, T. B. Saw, B. Ladoux, and C. T. Lim, Nature Communications8, 1517 (2017)

    2017

  12. [12]

    Glentis, C

    A. Glentis, C. Blanch-Mercader, L. Balasubramaniam, T. B. Saw, J. d’Alessandro, S. Janel, A. Douanier, B. Delaval, F. Lafont, C. T. Lim, D. Delacour, J. Prost, W. Xi, and B. Ladoux, Science Advances8, eabn5406 (2022)

    2022

  13. [13]

    W. Tang, A. Das, A. F. Pegoraro, Y. L. Han, J. Huang, D. A. Roberts, H. Yang, J. J. Fredberg, D. N. Kotton, D. Bi, and M. Guo, Nature Physics18, 1371 (2022)

    2022

  14. [14]

    Pieuchot, J

    L. Pieuchot, J. Marteau, A. Guignandon, T. D. San- tos, I. Brigaud, P.-F. Chauvy, T. Cloatre, A. Ponche, T. Petithory, P. Rougerie, M. Vassaux, J.-L. Milan, N. T. Wakhloo, A. Spangenberg, M. Bigerelle, and K. Anselme, Nature Communications9, 3995 (2018)

    2018

  15. [15]

    Brandst¨ atter, D

    T. Brandst¨ atter, D. B. Br¨ uckner, Y. L. Han, R. Alert, M. Guo, and C. P. Broedersz, Nature Communications 14, 1643 (2023)

    2023

  16. [16]

    Salbreux and F

    G. Salbreux and F. J¨ ulicher, Physical Review E96, 032404 (2017), 1701.09095

    Pith/arXiv arXiv 2017

  17. [17]

    R. G. Morris and M. Rao, Physical Review E100, 022413 (2019)

    2019

  18. [18]

    Torres-S´ anchez, M

    A. Torres-S´ anchez, M. K. Winter, and G. Salbreux, PLOS Computational Biology18, e1010762 (2022)

    2022

  19. [19]

    Salbreux, F

    G. Salbreux, F. J¨ ulicher, J. Prost, and A. Callan-Jones, Physical Review Research4, 033158 (2022), 2201.09251

    Pith/arXiv arXiv 2022

  20. [20]

    Bell, S.-Z

    S. Bell, S.-Z. Lin, J.-F. Rupprecht, and J. Prost, Physical Review Letters129, 118001 (2022), 2203.05644

    Pith/arXiv arXiv 2022

  21. [21]

    Sknepnek and S

    R. Sknepnek and S. Henkes, Physical Review E91, 022306 (2015), 1407.8516

    Pith/arXiv arXiv 2015

  22. [22]

    Ehrig, J

    S. Ehrig, J. Ferracci, R. Weinkamer, and J. W. C. Dun- lop, Physical Review E95, 062609 (2017), 1610.05987

    Pith/arXiv arXiv 2017

  23. [23]

    Alaimo, C

    F. Alaimo, C. K¨ ohler, and A. Voigt, Scientific Reports7, 5211 (2017)

    2017

  24. [24]

    Henkes, M

    S. Henkes, M. C. Marchetti, and R. Sknepnek, Physical Review E97, 042605 (2018), 1705.05166

    Pith/arXiv arXiv 2018

  25. [25]

    Ai, B.-y

    B.-q. Ai, B.-y. Zhou, and X.-m. Zhang, Soft Matter16, 4710 (2020)

    2020

  26. [26]

    Hindes, V

    J. Hindes, V. Edwards, S. Kamimoto, G. Stantchev, and I. B. Schwartz, Physical Review E102, 022212 (2020), 2007.15986

    Pith/arXiv arXiv 2020

  27. [27]

    Omori,Mechanics and Geometry of Fields (Japanese) (Shokabo, Tokyo, 1991)

    H. Omori,Mechanics and Geometry of Fields (Japanese) (Shokabo, Tokyo, 1991)

    1991

  28. [28]

    Frankel,The Geometry of Physics(Cambridge Uni- versity Press, 2011)

    T. Frankel,The Geometry of Physics(Cambridge Uni- versity Press, 2011)

    2011

  29. [29]

    Y. Fily, A. Baskaran, and M. F. Hagan, arXiv 10.48550/arxiv.1601.00324 (2016), 1601.00324

    Pith/arXiv arXiv doi:10.48550/arxiv.1601.00324 2016

  30. [30]

    E. D. Mackay, G. Janzen, D. A. M. Fernandez, and R. Sknepnek, arXiv 10.48550/arxiv.2505.24730 (2025), 2505.24730

    work page doi:10.48550/arxiv.2505.24730 2025

  31. [31]

    Janzen, E

    G. Janzen, E. D. Mackay, R. Sknepnek, and D. A. Matoz- Fernandez, arXiv 10.48550/arxiv.2507.23616 (2025), 2507.23616

    work page doi:10.48550/arxiv.2507.23616 2025

  32. [32]

    Schweitzer, W

    F. Schweitzer, W. Ebeling, and B. Tilch, Physical Review Letters80, 5044 (1998)

    1998

  33. [33]

    Ohta and T

    T. Ohta and T. Ohkuma, Physical Review Letters102, 154101 (2009)

    2009

  34. [34]

    Hiraiwa, A

    T. Hiraiwa, A. Nagamatsu, N. Akuzawa, M. Nishikawa, and T. Shibata, Physical Biology11, 056002 (2014)

    2014

  35. [35]

    Hiraiwa, Physical Review E99, 012614 (2019)

    T. Hiraiwa, Physical Review E99, 012614 (2019)

    2019

  36. [36]

    A. J. Roberts, arXiv 10.48550/arxiv.1210.0933 (2012), 1210.0933

    Pith/arXiv arXiv doi:10.48550/arxiv.1210.0933 2012

  37. [37]

    Brewin, Classical and Quantum Gravity26, 175017 (2009), 0903.2087

    L. Brewin, Classical and Quantum Gravity26, 175017 (2009), 0903.2087. 10 •Movie 1:Time evolution for Fig. 3(a) (K= 0.05,σ= 0.2). •Movie 2:Time evolution for Fig. 3(b) (K= 0.1,σ= 0.2). •Movie 3:Time evolution for Fig. 3(c) (K= 1.0,σ= 0.2). •Movie 4:Time evolution for Fig. 4(a) (oblate spheroid;e= 0.5,K= 0.2,σ= 0.2). •Movie 5:Time evolution for Fig. 4(b) (o...

    Pith/arXiv arXiv 2009