Pith. sign in

REVIEW 4 major objections 5 minor 20 references

Counting Equilibria of the Electrostatic Potential

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper proves a new upper bound on electrostatic equilibria, builds configurations whose equilibrium-per-charge ratio surpasses 25/7, and refutes a 2007 conjecture via the truncated octahedron.

desk verdict The Bezout upper bound is real; the record ratio and counterexample are not yet fully proven, so read carefully before trusting the headline claims. read the letter →

arxiv 2501.05315 v2 pith:UQTHFNPB submitted 2025-01-09 cs.CG math-phmath.COmath.MP

Herbert Edelsbrunner , Christopher Fillmore , Gonçalo Oliveira This is my paper · ORCID
classification cs.CGmath-phmath.COmath.MP
keywords electrostaticpotentialelectricfieldzeroesequilibriaMorsetheoryBezouttheoremVoronoitessellationtruncatedoctahedronMaxwellconjecture
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 studies the number of equilibria—points where the electric field vanishes—of a potential generated by n positive point charges in $R^{3}$. It establishes three things: the electric field has at most 2^n(3n−2)^3 isolated critical points when one charge is placed generically; iterating a square anti-prism construction gives configurations whose equilibrium-to-charge ratio exceeds 25/7−ε for any ε>0; and the 24 unit charges at the vertices of a truncated octahedron violate a 2007 conjecture of Gabrielov, Novikov and Shapiro, since V1 has 18 index-1 equilibria while the limiting distance function has only 14. The new upper bound is the best known to date, and the counterexample shows that the electrostatic potential can have more equilibria than the distance function defined by the same point charges. The results leave Maxwell's 1873 quadratic upper bound open while suggesting it is far from tight.

What carries the argument

The upper bound is carried by a polynomial reformulation: clearing denominators in $\nabla V=0$ and introducing variables $u_m$ with $u_m^2=\|x-A_m\|^2$ turns the critical-point equations into a system $P(x,u)=0$ whose degrees are $3n-2$ for the three gradient components and $2$ for the n radius equations. Affine Bezout then bounds isolated zeroes by the product $2^n(3n-2)^3$, and Lemma 2 shows the non-degenerate critical points are isolated zeroes when one charge is generic. The record ratio is built by iterated substitution: replacing each charge by a small copy of the same solid and assuming the equilibria simply add gives $m(n^\ell-1)/(n-1)$ equilibria for $\ell$ layers, and the square anti-prism attains $m/(n-1)=25/7$ in the limit. The counterexample is the truncated octahedron, whose symmetry lets the authors locate equilibria along intersections of reflection planes and associate them with facets and edges; the count is 18 index-1 saddles for $V_1$ versus 14 for the distance-function limit.

What would settle it

Track all equilibria numerically to high precision in the two-layer iterated square anti-prism: Eq. (9) predicts $m(n^2-1)/(n-1)=25\cdot 9=225$ equilibria, so any count different from 225 would falsify the additivity premise behind Result 2. Independently, perturb the 24 truncated-octahedron charges generically and recount index-1 equilibria of $V_1$: a drop from 18 to 14 would show the counterexample to Conjecture 2 rests on the non-generic symmetry.

Watch

Extended reading notes

Core claim

The central assertions are: (1) for any n and any charges, after moving a single chosen charge to a generic position, the isolated critical points of $V$ number at most $2^n (3n-2)^3$; (2) for every $\varepsilon>0$ there are unit-charge configurations with $k/n > 25/7 - \varepsilon$, obtained by layering iterated copies of a square anti-prism; and (3) the potential generated by unit charges at the vertices of the truncated octahedron has $18$ index-1 equilibria, whereas the limiting distance function $E(x)=\min_i \|x-A_i\|$ has only $14$ index-1 equilibria and the same $36$ index-2 equilibria, so the total for $V_1$ exceeds the total for $E$, contradicting Conjecture 1.8(a) of Gabrielov, Novikov and Shapiro. The paper also proves that, for even $p=2r$, the modified potential $V_p$ has at most $(r(n-1)+2-r)^3$ isolated critical points, and it reports numerical and analytic equilibrium counts for charges placed at vertices of Platonic, Archimedean, Catalan, prism and anti-prism solids.

Load-bearing premise

The record-ratio construction depends on the unproven additivity formula in Eq. (9): after replacing each charge by a tiny copy of the original solid, the total number of equilibria is assumed to be exactly $m(n^\ell-1)/(n-1)$, and no multi-scale perturbation argument is supplied.

Editorial extensions

If this is right

  • The previously known upper bound—roughly $5\cdot 9^{3+n}$—drops to $2^n(3n-2)^3$, though this is still far above Maxwell's conjectured $(n-1)^2$.
  • For the even-power potentials $V_{2r}$, the bound $(r(n-1)+2-r)^3$ is polynomial in $n$ for fixed $r$, with $V_2$ at most $n^3$ isolated critical points.
  • Conjecture 1.8(a) of Gabrielov, Novikov and Shapiro is false as stated: the electrostatic potential of equal unit charges can have more equilibria than the distance function of the same points.
  • The iterative anti-prism construction produces equilibrium-to-charge ratios approaching $25/7\approx 3.57$, the highest ratio found so far, from a base ratio of $25/8$ for the square anti-prism.
  • For $p>1$, the potentials $V_p$ have no maxima when all charges are positive, and the octahedron example shows that minima, forbidden for the harmonic case $p=1$, can appear for larger $p$.

Reading between the lines

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

  • A rigorous multi-scale perturbation proof for the additivity in Eq. (9) would promote the 25/7 record from construction to theorem; the paper states the formula without that proof.
  • The truncated octahedron is not generic, so whether the failure of Conjecture 2 persists under the generic-position hypothesis remains open; a small generic perturbation test would settle it.
  • One testable extension is to iterate other Archimedean solids and compare limiting ratios; the paper's census shows several candidates with per-vertex ratios above 3, but only the square anti-prism is iterated.
  • The layered construction suggests that high ratios can be amplified by substitution; if a similar additivity held for other solids, the 25/7 record might be a starting point rather than a ceiling.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The manuscript studies the number of critical points (equilibria) of the electrostatic potential of n point charges in R^3. It claims three main results: (1) an upper bound of 2^n(3n-2)^3 on the number of isolated critical points for generic position of one charge, obtained by clearing denominators and applying Bezout's theorem; (2) a family of iterated anti-prism configurations with equilibrium-to-charge ratio exceeding 25/7 - epsilon; and (3) a counterexample to Conjecture 1.8(a) of Gabrielov, Novikov and Shapiro, based on 24 unit charges at the vertices of a truncated octahedron. The paper also surveys equilibria for Platonic, Archimedean and Catalan solids, discusses the one-parameter family V_p, and formulates several conjectures about slices of Voronoi tessellations.

Significance. If fully established, the improved upper bound would be a substantial advance over the previous Thom-Milnor bound, and the Bezout-based clearing-of-denominators idea is attractive. The ratio construction would give a new lower-bound record, and a valid counterexample to the GNS conjecture would be an important result. However, in the present version the counterexample is not a counterexample to the conjecture as stated, the record ratio rests on an unproved additivity formula, and the proof of the upper bound has a nontrivial gap. The systematic polyhedral tables and the local-homology observations are useful computational and structural contributions.

major comments (4)
  1. [Section 2.3, Lemma 2] The proof that the Jacobian of the polynomial map P has full rank is incomplete. The argument shows only that the upper-left block dR/dx is nonsingular when the displayed sum is nonzero. Because the lower-right block dQ/du is invertible, full rank of the full Jacobian requires invertibility of the Schur complement dR/dx - (dR/du)(dQ/du)^{-1}(dQ/dx), which is never established. In addition, the displayed formula for dR_j/du_q omits the factor (x_j - p_{ij}); with that factor restored, the Schur complement is different and the claimed conclusion is not immediate. Thus the correspondence between non-degenerate equilibria and isolated non-degenerate zeroes of P, and hence the Bezout bound of Result 1, is not proven as written.
  2. [Section 2.5, Proposition 1] The degree computation in Proposition 1 is incorrect. For p = 2r, each factor r_m^{p+2} = r_m^{2r+2} = (||x - A_m||^2)^{r+1} has degree 2(r+1), not r-1. The numerator after clearing denominators therefore has degree 1 + 2(r+1)(n-1), not 1 + (n-1)(r-1). Consequently the stated bound (r(n-1)+2-r)^3, and in particular the n^3 bound for V_2, does not follow from the argument. For n = 2 and p = 2, the cleared numerator already has degree 5, showing the discrepancy directly.
  3. [Section 3.4, Eq. (9)] The additivity formula m(n^L - 1)/(n - 1) for the iterated construction is asserted without proof. A rigorous multi-scale argument must show that, at sufficiently small scale, every equilibrium of each layer persists, that the layers do not interact to create additional equilibria, and that the equilibria of the outer configuration survive the replacement of a charge by a tiny cluster. No such perturbation or convergence argument is supplied. Moreover, the base value m = 25 for the square anti-prism is reported from experiments in Section 3.3; Appendix C proves existence of some equilibria but not an exhaustive count. Therefore Result 2 is not established.
  4. [Section 4.2, Result 3] The truncated octahedron configuration is not in generic position: the center is a degenerate equilibrium, which is exactly the case excluded by Conjecture 2. To refute the conjecture one must exhibit a generic configuration, or at least prove that a small generic perturbation of the truncated octahedron preserves the inequality #1(V1) > #1(E). No such argument is given. The asserted counts of 18 index-1 and 36 index-2 equilibria are justified by symmetry-adapted observations and not by a complete analytic proof or certified computation. As presented, the example does not contradict the conjecture as stated.
minor comments (5)
  1. [Section 2.2, Theorem 1] The statement of Theorem 1 contains a typo: 'd1, ..., dd' should be 'd1, ..., dm'. The theorem should also state the standard hypotheses under which the affine Bezout bound applies to non-degenerate zeros.
  2. [Section 2.3, Lemma 2] The formula for dQ_m/dx_s contains a typo: 'xr - p_{ms}' should be 'x_s - p_{ms}'. The displayed formula for dR_j/du_q is also missing the factor (x_j - p_{ij}).
  3. [Section 3.3] The phrase 'Our experiments show' should be replaced by a clear distinction between proven results and numerical evidence; Appendix C proves the existence of some equilibria but not the exhaustive counts used in Eq. (8).
  4. [Appendix C, Theorem 5] There is a typo in the proof: 'if h < sqrt(2)h' should read 'if h < sqrt(2)R', and 'x = 2 = 0 = x3' should read 'x2 = 0 = x3'.
  5. [Section 6] The question about exceeding 25n/7 should be phrased conditionally on the validity of Result 2, since the claimed record depends on the unproved additivity formula in Eq. (9).

Circularity Check

0 steps flagged · score 1.0 of 10

No circular reduction found; the unproven additivity formula and generic-position gap are soundness concerns, not circularity.

full rationale

Walking the derivation chain: Result 1 (Section 2) is a self-contained Bezout count: Lemma 1 rewrites grad V = 0 as common zeros of polynomials R_j, the u-variables make the system polynomial with degrees 3n-2 and 2, Lemma 2 (proved in-paper for generic p1) identifies non-degenerate critical points with isolated zeros, and affine Bezout then gives (3n-2)^3 * 2^n. No parameter is fitted and no input is renamed as an output. Result 2 (Section 3.4) rests on Eq. (9), 'm * (n^{ell-1} + n^{ell-2} + ... + 1) = m/(n-1) * (n^ell - 1)', which assumes without proof that each small copy contributes m equilibria and outer equilibria persist; this is an unproven multi-scale additivity assumption, hence a soundness gap, but not a definitional equivalence or a fitted quantity. Result 3 (Section 4.2) asserts 'E has 14 1-saddles and 36 2-saddles, which we compare to the 18 1-saddles and 36 2-saddles of V', based on symmetry-adapted observation and no full generic perturbation argument; Conjecture 2 is restricted to 'generic position', and the truncated octahedron has a degenerate center, so the example may be outside the conjecture's scope. That is a correctness concern, not circularity. The self-citations ([5], [2], [14]) are peripheral: [5] provides a Voronoi lower-bound theorem in the discussion, [2] motivates a question, and [14] is an application pointer; none is load-bearing for the main results. No equation reduces by construction to its own input, and no 'prediction' is a renamed fit. Score 1 reflects the absence of circularity while noting the two unproven jumps.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new entities. It relies on standard theorems plus two unproven structural assumptions: the layer additivity in the iterated construction and the exactness of the computational equilibrium counts. The relative height and layer scale are free parameters tuned by hand to make the construction work.

free parameters (2)
  • relative height h/R of the square anti-prism = not specified; chosen in a range (h/R < sqrt(2)) to maximize equilibria
    The record construction requires a square anti-prism whose center is a non-degenerate 1-saddle and which has one 2-saddle per edge and one 1-saddle per triangle; the exact height is not given, only its existence.
  • scale factor between layers in the iterated construction = not specified; 'much smaller copy'
    For the additivity formula to hold, each successive copy must be small enough; no quantitative bound is provided.
assumptions (5)
  • standard math Affine Bezout's theorem bounds the number of non-degenerate zeroes of a polynomial system by the product of degrees.
    Used in Section 2.2 to prove Result 1.
  • standard math Morse theory and the Euler-Poincare relation for S^3 imply m2 - m1 = n - 1 for the compactified potential.
    Used in Section 3.1 to derive the lower bound and structure.
  • domain assumption GNS Theorem 1.7: for large p, equilibria of Vp correspond to effective Voronoi cells, giving #1 = 14 for the truncated octahedron.
    External theorem used in Section 4.2 to define the target #1.
  • ad hoc to paper In the iterated construction, the number of equilibria of the multi-scale configuration is the sum of the equilibria of each layer (Eq. 9).
    Unproven perturbation assumption central to Result 2.
  • ad hoc to paper The counts of 18 index-1 and 36 index-2 equilibria for the truncated octahedron are correct.
    Reported from numerical observation in Section 4.2 without a complete proof or code.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Counting Equilibria of the Electrostatic Potential." pith.science (2026). https://pith.science/paper/UQTHFNPB

@misc{pith2026250105315,
  author       = {Pith},
  title        = {Pith review of: Counting Equilibria of the Electrostatic Potential},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UQTHFNPB}},
  note         = {Machine review of arXiv:2501.05315}
}
abstract

In 1873, James C. Maxwell conjectured that the electric field generated by $n$ point charges in generic position has at most $(n-1)^2$ isolated zeroes. The first (non-optimal) upper bound was only obtained in 2007 by Gabrielov, Novikov and Shapiro, who also posed two additional interesting conjectures. In this article, we give the best upper bound known to date on the number of zeroes of the electric field, and construct a counterexample to a conjecture of Gabrielov, Novikov and Shapiro that the number of equilibria cannot exceed those of the distance function defined by the unit point charges. Finally, we note that it is quite possible that Maxwell's quadratic upper bound is not tight, so it is prudent to find smaller bounds. Hence, we also explore examples and construct configurations of charges achieving the highest ratios of the number of electric field zeroes by point charges found to this day.

Figures

Figures reproduced from arXiv: 2501.05315 by the authors.

Figure 1
Figure 1. Upper row, from left to right: the binary functions on the unit 2-sphere for a non￾critical point, a minimum, a 1-saddle, a 2-saddle, and a maximum. Lower row, from left to right: the binary functions for the centers of the tetrahedron, cube, octahedron, dodecahedron, and isoc￾ahedron (these are degenerate equilibria and so are neither 1-saddles nor 2-saddles). W ⊆ S 2 for the white points, the local homology of x i… view at source ↗
Figure 2
Figure 2. From left to right: the hexagonal, pentagonal, square anti-prisms with the heights chosen to maximize the number of equilibria. The ratios of equilibria over vertices are 37/12 < 31/10 < 25/8, respectively. Observe how a ring of alternating 1- and 2-saddles gets successively more concentrated around the center. In addition to the results on the maximum number of equilibria for anti-prims, we discover an interesting … view at source ↗
Figure 3
Figure 3. Cut-away views of three level sets of V1 (upper row) and three level sets of V1.3 (lower row) defined by point sources at the vertices of the octahedron. From left to right: the values are chosen slightly less than, equal to, and slightly greater than the potential at the center of the octahedron. Removing the front of the surface reveals some of the complication at the center, which for V1 is a degenerate equilibri… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Equilibria of the electrostatic potential generated by unit point charges at the vertices of the truncated octahedron. In total there are 36 light blue 2-saddles, 18 dark red 1-saddles, and the degenerate equilibrium at the center. For better visualization, we split th…
Figure 5
Figure 5. Figure 5: The equilibria of V2 generated by unit point charges at the vertices of the truncated octahedron. Compared with V1, we note a drastically reduced number of 1-saddles and a minimum at the origin; see [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]
Figure 6
Figure 6. Figure 6: Plot of f ′ (x) whose zeroes correspond to electrostatic points lying in the line P1 ∩ P2. particular, that such points lie along the straight lines passing through the cube’s center and the midpoints in its edges. These two points are in the same orbit of the group of…
Figure 7
Figure 7. Figure 7: Left: the Rhombicuboctahedron and the equilibria of the electrostatic potential gen￾erated by unit point charges at its vertices. There are 36 light blue 2-saddles, 8 dark red 1-saddles, and a degenerate equilibrium at the center. Right: the Elongated Square Gyrobicupo…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

20 extracted references · 19 canonical work pages

  1. [1]

    Aurenhammer and H

    F. Aurenhammer and H. Edelsbrunner.An optimal algorithm for constructing the weighted Voronoi diagram in the plane. Pattern Recognition 17 (1984), 251–257

    work page 1984

  2. [2]

    Proc. 13th Internat. Conf. Comput. Vision, 2011

    C. Chen and H. Edelsbrunner. Diffusion runs low on persistence fast. In “Proc. 13th Internat. Conf. Comput. Vision, 2011”, 423–430

    work page 2011

  3. [3]

    Delaunay

    B. Delaunay. Sur la sph` ere vide.Izv. Akad. Nauk SSSR, Otdelenie Matematicheskii i Estestvennykh Nauk 7, 1934, 793–800

    work page 1934

  4. [4]

    Gao and M

    P. Gao and M. R. Douglas Geodesics on Calabi-Yau manifolds and winding states in non-linear sigma models. Frontiers in Physics 1, 2013, 26

    work page 2013

  5. [5]

    Proc. 40th Internat. Sympos. Comput. Geom., 2024

    H. Edelsbrunner and J. Pach. Maximum Betti numbers of ˇCech complexes. In “Proc. 40th Internat. Sympos. Comput. Geom., 2024”, 53:1–53:14. 2Indeed, for the weighted squared distance in (12), each point generates at most one interval along the line. This is not true for the weighted squared distance in (13), but it is not difficult to see that the point wit...

    work page 2024

  6. [6]

    Gabrielov, D

    A. Gabrielov, D. Novikov and B. Shapiro. Mystery of point charges. Proc. London Math. Soc. 95 (2007), 443–472

    work page 2007

  7. [7]

    A remark on Maxwell’s conjecture for planar charges

    Killian, Kenneth. A remark on Maxwell’s conjecture for planar charges. Complex Variables and Elliptic Equations 54.12 (2009): 1073-1078

    work page 2009

  8. [8]

    Khovanski ˘ ı, Askold G.Fewnomials. Vol. 88. American Mathematical Society, 1991

    work page 1991

Show all 20 references
  1. [9]

    J.C. Maxwell. A Treatise on Electricity and Magnetism, vol. 1. Reproduction of 3-rd revised edition, Dover, New York, 1954

    1954

  2. [10]

    McMullen

    P. McMullen. On the upper-bound conjecture for convex polytopes. J. Combinat. Theory, Ser. B 10 (1971), 187–200

    1971

  3. [11]

    J. Milnor. Morse Theory. Princeton Univ. Press, Princeton, New Jersey, 1963

    1963

  4. [12]

    Morse and S

    M. Morse and S. Cairns. Critical Point Theory in Global Analysis and Differential Topology. Volume 33 in Pure and Applied Mathematics, Academic Press, New York and London, 1969

    1969

  5. [13]

    Okabe, B

    A. Okabe, B. Boots, K. Sugihara and S.N. Chiu. Spatial Tessellations: Concepts and Applications of Voronoi Diagrams. Second edition, Wiley, Hoboken, New Jersey, 2000

    2000

  6. [14]

    Oliveira Electrostatics and geodesics on K3 surfaces

    G. Oliveira Electrostatics and geodesics on K3 surfaces. arXiv preprint 2302.08354 2023

    2023 arXiv

  7. [15]

    G. Voronoi. Nouvelles applications des param` etres continus ` a la th´ eorie des formes quadratiques.J. Reine Angew. Math. 133 (1907), 97–178 and 134 (1908), 198–287 and 136 (1909), 67–182

    1907

  8. [16]

    Polynomial automorphisms of Cn

    Rusek, Kamil, and Tadeusz Winiarski. Polynomial automorphisms of Cn. Univ. Iagell. Acta Math 24.1 (1984): 984

    1984

  9. [17]

    Upper bounds for the number of isolated critical points via the Thom–Milnor theorem

    Zolotov, Vladimir. Upper bounds for the number of isolated critical points via the Thom–Milnor theorem. Analysis and Mathematical Physics 13.5 (2023): 81. 20 HERBERT EDELSBRUNNER, CHRISTOPHER FILLMORE, AND GONC ¸ ALO OLIVEIRA Appendix A. Analytic Proof of Existence of Equilibr...

    2023

  10. [18]

    Hence, there exists ε >0, such that ∂V ∂x3 (0, 0, t) is positive for t ∈ (0, ε) and negative for t ∈ (−ε, 0)

    From this it follows that ∂V ∂x3 (0, 0, x3) = − XN i=1 x3 + h 2 r3 i (0, 0, x3 + h 2 ) − XN i=1 x3 − h 2 r3 i (0, 0, x3 − h 2 )(41) = −N x3 + h 2 R2 + (x3 + h 2 )2 3/2 − N x3 − h 2 R2 + (x3 − h 2 )2 3/2 .(42) Evaluating this at the top and bottom of the β-rotated prism, we obt...

  11. [19]

    COUNTING EQUILIBRIA OF THE ELECTROSTATIC POTENTIAL 27 Proof

    Then, there are electrostatic points located along the line segments connecting the center of the anti-prism to the midpoints of its lateral edges. COUNTING EQUILIBRIA OF THE ELECTROSTATIC POTENTIAL 27 Proof. We shall consider the case of the line segment connecting the center...

  12. [20]

    ,(45) with terms of order exceeding 4 not shown

    − 14 81 x4 3 − 7 54 x2 1x2 2 + 14 27 x2 3(x2 1 + x2 2) + . . . ,(45) with terms of order exceeding 4 not shown. Again, we compute the Taylor expansion at the origin as a way to understand the degenerate critical point there for h = √ 2R. Lemma 7. Let V : R3 → R be the electros...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.