On the high density behavior of Hamming codes with fixed minimum distance

F.Zamponi; G.Parisi

REVIEW

On the high density behavior of Hamming codes with fixed minimum distance

G.Parisi, F.Zamponi

Statistical Mechanics · Submitted 28 Feb 2006 · 19 references

Pith's verdict: A serious, honest heuristic that maps a liquid-state instability onto the Varshamov-Gilbert bound, but whose central sign of the entropy rests on unproven HNC closure.

T0 · AI review

How this review was made: An AI referee (deepseek-v4-flash) read the paper on August 28, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

T0 AI review An AI read the paper and answered a fixed list of questions.
T1 Audit Automatic checks that the references and files the paper relies on exist and match.
T2 Compute The paper's code or calculations re-run with the same results.
T3 Formal statement The claims written out in a formal math language a computer can read.
T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 4 major objections · 4 minor · Ask about this paper · The compiled record · Read the full review →

Challenge this review Comment (0) Read on arXiv ↗

Cite ▾

arXiv IDcond-mat/0602661 (version 1)

Pith IDYIFKXFS7, Pith's permanent record of this paper and its review

CitationF.Zamponi, G.Parisi. "On the high density behavior of Hamming codes with fixed minimum distance." arXiv:cond-mat/0602661 (2006). https://pith.science/paper/YIFKXFS7

Cite this reviewBibTeX for Pith's review of this paper

Claim paper

Why Pith says this

Highlight any text to comment on it.

Editor's note

read the letter

Parisi and Zamponi have written a serious heuristic paper. They take the Frisch-Percus large-dimensional liquid idea from the continuum, adapt it to the Hamming hypercube, work through the Krawtchouk/Fourier asymptotics in enough detail to follow, and end with a concrete scenario: within HNC, the liquid entropy goes negative at phi of order log n/n, which is exponentially close to the Varshamov-Gilbert bound. They then conjecture a phase transition, most likely a glass transition, and observe that if no other phase intervenes, VG would be asymptotically tight. That is a big 'if,' but the paper gives a physical mechanism and a specific density scale where something has to break.

What is new here is the discrete extension itself. The saddle-point treatment of the Krawtchouk transform, the step-function solution with exponentially small corrections, and the closed form for the entropy are done transparently. The even-d sublattice instability is a nice phenomenological observation, with some Monte Carlo support. The numerical check at n=36 is limited but consistent with the asymptotic formulas.

The soft spots are real, and the stress-test note lands on the right one. The entropy sign is computed from the HNC free energy, and the paper assumes that any resummation containing ring diagrams is equivalent at large n, importing that from the continuum. At phi ~ log n/n, the terms that decide the sign, -n phi and -0.5 e^{n phi}, are both O(n), so an omitted non-ring contribution of the same order could flip the sign. The choice of h1 in Eq (50) is an ansatz; the min rule around Eq (41) is plausible but not derived from the contour deformation; and the odd-d/even-i symmetry trick is delicate. These are acknowledged limitations, not hidden flaws. The paper is explicit that the results are not conclusive.

Credit where due: the authors do not oversell. The self-citations to their continuum work are used as analogy, not as circular support. The derivations are laid out so a reader can see exactly where the approximation enters.

Bottom line: this is a paper for a reader who wants a physics-informed conjecture on an open problem in coding theory, not for a reader who needs a proof. It deserves a serious referee. I would accept it for peer review. The central conjecture may well be wrong, but the paper is honest, novel, and likely to provoke follow-up work. A revision that quantifies the sensitivity to non-ring corrections would strengthen it considerably.

Referee report

Major objections (4)

[Section III, after Eq. (12), and Section III C] The central conclusion that the liquid must become unstable at φ∼log n/n is obtained from Eq. (62), which is the HNC free energy (23) evaluated on the constructed solution. The equivalence of HNC to the exact liquid is asserted rather than proved ('any resummation will be equivalent...'), and at the relevant density the two terms controlling the sign of S, -nφ and -(1/2)e^{nφ}, are both O(n); omitted non-ring diagrams, or even a 1/n correction in the even-i projection discussed in Appendix A, can change the sign. The paper should either estimate the omitted diagrams at this density or prove a bound, and should state the instability as conditional on the HNC approximation rather than as a property of the exact liquid.
[Section III B, Eq. (50)] The function h1(k) in Eq. (50) is chosen as the simplest solution of an inequality (49), but no argument is given that it is the solution selected by the HNC equations (24), nor that it is the relevant stationary point of (23). The subsequent Y consistency equation (56) only checks exponential smallness of the chosen form; it does not check the stationarity condition. Without a derivation from the HNC equations, the constructed h(x), and hence the entropy (62), may describe an admissible correlation function but not the liquid state.
[Section III A, Eq. (41), and Appendix A] The 'min Re' rule for selecting the saddle point in Eq. (41) is stated from geometric considerations but not derived, and in region A the phase is complex so cancellation effects are not controlled. In addition, the inversion in Eq. (55) is restricted to even i, with odd-i values obtained through h(i-1)=h(i); Appendix A says a factor 1/n is discarded. Since the claimed instability threshold is at φ∼(log n)/n and the entropy balance in Eq. (63) contains O(log n) terms, these 1/n and parity effects are not obviously negligible and need to be quantified.
[Section III D] The numerical check is restricted to n=36 and to the HNC equations themselves; it shows that the iterative solution approaches the proposed ansatz, but it does not test the key assumption that HNC is equivalent to the exact hard-sphere system at n→∞ and φ>0. A Monte Carlo computation of the liquid correlation function and entropy, or at least a comparison of different closures (for example PY versus HNC), would be needed to support the claim that the negative entropy is a real instability rather than an artifact of the closure.

Minor objections (4)

[Section III A, Eq. (35)] The contour C is described only as 'embracing the negative part of the real z axis'; a more precise specification would be needed to make the saddle-point analysis reproducible.
[Section III E] The statement 'S_glass(φ) ≤ S_liquid(φ)' is introduced via reference [15] with no derivation; it should be explicitly marked as a heuristic analogy.
[Figure 5] The caption does not specify all parameters used in the analytic lines, such as the value of n and d in Eq. (65), making it harder to judge the claimed agreement.
[Abstract] The title and abstract refer to 'Hamming codes', but the paper treats general binary error-correcting codes with fixed minimum distance; the terminology could be clarified in the introduction.

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

The liquid-state description of Hamming codes breaks down at densities near the Varshamov–Gilbert bound, predicting that bound is asymptotically tight.

The reading

This paper treats the maximum size of a binary error-correcting code as a statistical-mechanics problem: codewords are hard spheres on the $n$-dimensional binary hypercube $\mathbb{H}_2^n$, forbidden from being closer than Hamming distance $d$. It argues that the liquid phase, continued through the hypernetted-chain equations, remains a valid description only up to a reduced density $\tilde{\rho}=\rho V_d$ of order $n$, i.e. $\phi=(1/n)\ln\tilde{\rho}\sim(\ln n)/n$, extremely close to the Varshamov–Gilbert lower bound. At that point the HNC entropy of the liquid branch turns negative, which is impossible for a discrete system, so the paper concludes the liquid must become unstable and a phase transition must occur. If this holds, and no other phase beats the conjectured glass, the asymptotic rate of binary codes is exactly the Varshamov–Gilbert value $R(\delta)=1-H(\delta)$.

What carries the argument

The central object is the hard-core lattice gas on the binary hypercube with radius-$d$ exclusion, analyzed through the hypernetted-chain (HNC) approximation to the liquid equation of state. The argument runs on the Fourier transform on the hypercube, whose kernel is the Krawtchouk polynomial $F_n(a,i)$; in the $n\to\infty$ limit this transform is evaluated by saddle point, giving $\Phi(k,x)\sim \operatorname{Re} e^{n f(k,x)}$. The liquid correlation is the step function $g(i)=\theta(i-d+1)$, corrected by an exponentially small term built from the Fourier mode at a critical wavevector $k_c(\delta,\phi)$, determined by $f(k_c,\delta)=v(\delta)-\phi$. The HNC free energy then yields an entropy whose leading density dependence is $S\approx 1-n\phi+nv(\delta)-\tfrac12 e^{n\phi}$; the sign change of $S$ at $\phi\sim(\ln n)/n$ is the mechanism that forces the liquid instability.

Extended reading notes

Core claim

The central claim is that the liquid phase of hard spheres on $\mathbb{H}_2^n$, defined by the virial expansion and continued through the hypernetted-chain resummation, ceases to exist at a reduced density $\tilde{\rho}\sim n$, which is exponentially close to the Varshamov–Gilbert density. The authors construct an asymptotic solution to the liquid-state positivity conditions of the form $g(i)=\theta(i-d+1)[1+\exp(n\tilde{h}(i))]$, an exponentially small correction to the step function, and show this solution exists up to densities $\phi_{\rm HNC}(\delta)$ and $\phi_S(\delta)$. Evaluating the HNC free energy along this branch gives an entropy $S\approx 1-n\phi+nv(\delta)-\tfrac12\ln n-\tfrac12 e^{n\phi}$, which becomes negative for $\phi\sim(\ln n)/n$. Since a discrete system cannot have negative entropy, the paper infers that the liquid branch cannot represent the physical high-density state and that a phase transition must occur at or below that density. By analogy with hard spheres in large continuous dimension the authors conjecture a glass transition, and for even $d$ they present evidence for a first-order transition to a phase localized on one parity sublattice; absent other phases, the Varshamov–Gilbert bound would be asymptotically tight.

Load-bearing premise

The load-bearing premise is that the hypernetted-chain approximation, or any resummation containing the ring diagrams, correctly describes the physical liquid branch for $n\to\infty$ up to exponentially large densities, so that the negative entropy is a real instability of the true liquid rather than an artifact of the approximation.

Editorial extensions

If this is right

The liquid-state route to bounding $A(n,d)$ has a built-in ceiling: no liquid correlation function of the HNC type can represent the code above $\tilde{\rho}\sim n$, so improvements over the Varshamov–Gilbert lower bound must come from non-liquid phases or explicit constructions.
If no crystalline or sublattice phase beats the glass for odd $d$, the asymptotic rate equals the Varshamov–Gilbert value $R(\delta)=1-H(\delta)$, shrinking the gap between the known lower and upper bounds.
For even $d$, the predicted first-order transition to a phase localized on one parity sublattice may set the maximum density below the glass threshold, so parity should be treated as a separate case in asymptotic bounds.
The scaling $\phi\sim(\ln n)/n$ places the transition region extremely close to the Varshamov–Gilbert density, meaning finite-size numerical tests will need $n$ large enough to resolve the logarithmic shift.
The conjectured glass phase, if it exists, would have entropy below the liquid entropy and would disappear at a density of the same order, making the maximum density scale as $\tilde{\rho}\sim n$.

Reading between the lines

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

If the negative-entropy criterion is taken as an upper-bound principle, one could try to turn it into a rigorous proof that any admissible correlation function has nonnegative entropy only for $\phi\lesssim(\ln n)/n$, which would make the Varshamov–Gilbert bound asymptotically tight from the liquid side.
The same saddle-point Fourier analysis could be rerun for $q$-ary Hamming spaces or constant-weight codes, where analogous hard-core liquids may show entropy crises at similar near-bound densities.
The parity sublattice transition for even $d$ suggests explicit code families built on one parity class may already saturate or exceed the VG density at finite $n$; testing whether such families beat VG would distinguish the two transition scenarios.
A replicated free-energy computation for this discrete model, along the lines used for continuum hard spheres, would either confirm the glass transition or show the analogy breaks down.

Original abstract

We discuss the high density behavior of a system of hard spheres of diameter d on the hypercubic lattice of dimension n, in the limit n -> oo, d -> oo, d/n=delta. The problem is relevant for coding theory. We find a solution to the equations describing the liquid up to very large values of the density, but we show that this solution gives a negative entropy for the liquid phase when the density is large enough. We then conjecture that a phase transition towards a different phase might take place, and we discuss possible scenarios for this transition. Finally we discuss the relation between our results and known rigorous bounds on the maximal density of the system.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13-20260917 via OpenRouter, question bank abstract-p1-20260923, 2026-09-23: 1 request, 4,210 input tokens, $0.0002 reported.

What the abstract itself states, as probabilities: claims first 0.03 · states limitation 0.94 · human participants 0.01 · controlled comparison 0.03 · randomized assignment 0.01 · states quantitative result 0.15 · compares with prior results 0.83 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

We discuss the high density behavior of a system of hard spheres of diameter d on the hypercubic lattice of dimension n, in the limit n -> oo, d -> oo, d/n=delta. The problem is relevant for coding theory. We find a solution to the equations describing the liquid up to very large values of the density, but we show that this solution gives a negative entropy for the liquid phase when the density is large enough. We then conjecture that a phase transition towards a different phase might take place, and we discuss possible scenarios for this transition. Finally we discuss the relation between our results and known rigorous bounds on the maximal density of the system.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	We discuss the high density behavior of a system of hard spheres of diameter d on the hypercubic lattice of dimension n, in the limit n -> oo, d -> oo, d/n=delta.	background 0.52	0.32	0.04	asserted 0.66
s1	The problem is relevant for coding theory.	background 0.97	0.05	0.02	asserted 1.00
s2	We find a solution to the equations describing the liquid up to very large values of the density, but we show that this solution gives a negative entropy for the liquid phase when the density is large enough.	result 0.98	0.96	0.38	asserted 1.00
s3	We then conjecture that a phase transition towards a different phase might take place, and we discuss possible scenarios for this transition.	interpretation 0.81	0.59	0.22	speculative 1.00
s4	Finally we discuss the relation between our results and known rigorous bounds on the maximal density of the system.	interpretation 0.75	0.46	0.07	asserted 0.69

What kind of paper the classifier reports this is: theoretical result 1.00 · evidence from proof or derivation 0.87.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures

**Figure 2.** Then Fn(nk, nx) ∼ e nσ(ζ−) ⇒ f(k, x) = σ(ζ−) . (40)

**Figure 5.** FIG. 5: Comparison of the numerical solution of the HNC equat

**Figure 6.** FIG. 6: Left: correlation function

Figures reproduced from arXiv:cond-mat/0602661 by the authors.

References

Works cited (19)

[1]

In the region ( k − 1 2 ) 2 + ( x − 1 2 ) 2 − 1 4 ≤ 0 (region A in Fig. 2) the solutions are complex, and ζ+ = ζ∗ −, σ (ζ+) = [ σ (ζ−)]∗. Thus one has Fn(nk, nx ) ∼ enσ (ζ− ) + enσ (ζ+) = Re enσ (ζ− ) ∼ enRe σ (ζ− ) cos[nIm σ (ζ−)] ⇒ f (k, x ) = σ (ζ−) . (39)

View on Pith
[2]

In region B of Fig. 2 the saddle points are real and positive with 0 ≤ ζ− ≤ ζ+, and σ (ζ±) are also real with σ (ζ+) > σ (ζ−) > 0. The point ζ− is then the closest saddle point to the original integration contour; moreover, it is a local mimimum of σ (ζ) along the real axis, so it is a maximum of σ (ζ) along the imaginary direction and the integration pat...

View on Pith
[3]

The behavior in the regions C,D,E can be obtained using the symmetr ies (18), (19), (20). Alternatively one can always choose the closest saddle point to the integration path that is a local minimum on the real axis: it turns out that one has to choose ζ− in the region E and ζ+ in the regions C and D. With this choice the symmetries (18), (19), (20) are r...

View on Pith 2002
[4]

A. Procacci and B. Scoppola, Statistical mechanics approach to coding theory , J. Stat. Phys. 96, 907–912 (1999)

View on Pith 1999
[5]

E. N. Gilbert, A comparison of signalling alphabets , Bell Syst. Tech. Jnl. 31, 504–522 (1952); R. R. Varshamov, Estimate of the number of signals in error correcting codes , Dokl. Akad. Nauk SSSR 117, 739–741 (1957)

View on Pith 1952
[6]

P. Delsarte, An algebraic approach to the association schemes of coding t heory, Philips Res. Rep. Suppl. 10, 1–97 (1973)

View on Pith 1973
[7]

R. J. McEliece, E. R. Rodemich, H. Rumsey and L. R. Welch, New upper bounds on the rate of a code via the Delsarte- MacWilliams inequalities , IEEE Trans.Inform.Theory 23, 157–166 (1977)

View on Pith 1977
[8]

G. Parisi and F. Slanina, A toy model for the mean-ﬁeld theory of hard-sphere liquids , Phys. Rev. E 62, 6554–6559 (2000)

View on Pith 2000
[9]

for the problem in the continuum - for this problem; but we believe t hat the computation is feasible and we leave it for future work. Clearly other phases may exist at least for some special values of n, d. For instance, we have some numerical evidence, for even d, of a ﬁrst-order phase transition toward a phase in which particles only occupy a subspace ...

View on Pith
[10]

A. Barg and D. B. Jaﬀe, Numerical results on the asymptotic rate of binary codes , in “Codes and Association Schemes” (A. Barg and S. Litsyn, Eds.), Amer. Math. Soc., Providence, 2001

View on Pith 2001
[11]

A. Samorodnitsky, On the Optimum of Delsarte’s Linear Program , Journal of Combinatorial Theory, Series A 96, 261–287 (2001)

View on Pith 2001
[12]

H. L. Frisch and J. K. Percus, High dimensionality as an organizing device for classical ﬂ uids, Phys. Rev. E 60, 2942–2948 (1999)

View on Pith 1999
[13]

G. Parisi and F. Zamponi, Amorphous packings of hard spheres in large space dimension , cond-mat/0601573

View on Pith Pith review arXiv
[14]

O. U. Uche, F. H. Stillinger and S. Torquato, On the realizability of pair correlation functions , Physica A 360, 21–36 (2006); S. Torquato and F. H. Stillinger, New provisional lower bounds on the optimal density of spher e packings , submitted

View on Pith 2006
[15]

J.-P. Hansen and I. R. MacDonald, Theory of simple liquids (Academic Press, London, 1986)

View on Pith 1986
[16]

Eric W. Weisstein, Krawtchouk Polynomial , From MathWorld–A Wolfram Web Resource, http://mathworld.wolfram.com/KrawtchoukPolynomial.html

View on Pith
[17]

A. Lenard, States of classical statistical mechanics of inﬁnitely man y particles: I, II , Arch. Rational. Mech. Anal. 59, 219–239, 241–256 (1975)

View on Pith 1975
[18]

T. Kuna, J. L. Lebowitz and E. Speer, On the realizability of point processes with speciﬁed one an d two particle densities , in Math.Frosch.Oberwolfach 43 (2004), Large Scale Stochastic Dynamics, ed. C. Landim, S. O lla, and H. Spohn; O. Costin and J. L. Lebowitz, On the construction of particle distributions with speciﬁe d single and pair densities , J....

View on Pith 2004
[19]

G. Parisi and F. Zamponi, The ideal glass transition of hard spheres , J. Chem. Phys. 123, 144501 (2005)

View on Pith 2005

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.

[1] [1]

In the region ( k − 1 2 ) 2 + ( x − 1 2 ) 2 − 1 4 ≤ 0 (region A in Fig. 2) the solutions are complex, and ζ+ = ζ∗ −, σ (ζ+) = [ σ (ζ−)]∗. Thus one has Fn(nk, nx ) ∼ enσ (ζ− ) + enσ (ζ+) = Re enσ (ζ− ) ∼ enRe σ (ζ− ) cos[nIm σ (ζ−)] ⇒ f (k, x ) = σ (ζ−) . (39)

View on Pith

[2] [2]

In region B of Fig. 2 the saddle points are real and positive with 0 ≤ ζ− ≤ ζ+, and σ (ζ±) are also real with σ (ζ+) > σ (ζ−) > 0. The point ζ− is then the closest saddle point to the original integration contour; moreover, it is a local mimimum of σ (ζ) along the real axis, so it is a maximum of σ (ζ) along the imaginary direction and the integration pat...

View on Pith

[3] [3]

The behavior in the regions C,D,E can be obtained using the symmetr ies (18), (19), (20). Alternatively one can always choose the closest saddle point to the integration path that is a local minimum on the real axis: it turns out that one has to choose ζ− in the region E and ζ+ in the regions C and D. With this choice the symmetries (18), (19), (20) are r...

View on Pith 2002

[4] [4]

A. Procacci and B. Scoppola, Statistical mechanics approach to coding theory , J. Stat. Phys. 96, 907–912 (1999)

View on Pith 1999

[5] [5]

E. N. Gilbert, A comparison of signalling alphabets , Bell Syst. Tech. Jnl. 31, 504–522 (1952); R. R. Varshamov, Estimate of the number of signals in error correcting codes , Dokl. Akad. Nauk SSSR 117, 739–741 (1957)

View on Pith 1952

[6] [6]

P. Delsarte, An algebraic approach to the association schemes of coding t heory, Philips Res. Rep. Suppl. 10, 1–97 (1973)

View on Pith 1973

[7] [7]

R. J. McEliece, E. R. Rodemich, H. Rumsey and L. R. Welch, New upper bounds on the rate of a code via the Delsarte- MacWilliams inequalities , IEEE Trans.Inform.Theory 23, 157–166 (1977)

View on Pith 1977

[8] [8]

G. Parisi and F. Slanina, A toy model for the mean-ﬁeld theory of hard-sphere liquids , Phys. Rev. E 62, 6554–6559 (2000)

View on Pith 2000

[9] [9]

for the problem in the continuum - for this problem; but we believe t hat the computation is feasible and we leave it for future work. Clearly other phases may exist at least for some special values of n, d. For instance, we have some numerical evidence, for even d, of a ﬁrst-order phase transition toward a phase in which particles only occupy a subspace ...

View on Pith

[10] [10]

A. Barg and D. B. Jaﬀe, Numerical results on the asymptotic rate of binary codes , in “Codes and Association Schemes” (A. Barg and S. Litsyn, Eds.), Amer. Math. Soc., Providence, 2001

View on Pith 2001

[11] [11]

A. Samorodnitsky, On the Optimum of Delsarte’s Linear Program , Journal of Combinatorial Theory, Series A 96, 261–287 (2001)

View on Pith 2001

[12] [12]

H. L. Frisch and J. K. Percus, High dimensionality as an organizing device for classical ﬂ uids, Phys. Rev. E 60, 2942–2948 (1999)

View on Pith 1999

[13] [13]

G. Parisi and F. Zamponi, Amorphous packings of hard spheres in large space dimension , cond-mat/0601573

View on Pith Pith review arXiv

[14] [14]

O. U. Uche, F. H. Stillinger and S. Torquato, On the realizability of pair correlation functions , Physica A 360, 21–36 (2006); S. Torquato and F. H. Stillinger, New provisional lower bounds on the optimal density of spher e packings , submitted

View on Pith 2006

[15] [15]

J.-P. Hansen and I. R. MacDonald, Theory of simple liquids (Academic Press, London, 1986)

View on Pith 1986

[16] [16]

Eric W. Weisstein, Krawtchouk Polynomial , From MathWorld–A Wolfram Web Resource, http://mathworld.wolfram.com/KrawtchoukPolynomial.html

View on Pith

[17] [17]

A. Lenard, States of classical statistical mechanics of inﬁnitely man y particles: I, II , Arch. Rational. Mech. Anal. 59, 219–239, 241–256 (1975)

View on Pith 1975

[18] [18]

T. Kuna, J. L. Lebowitz and E. Speer, On the realizability of point processes with speciﬁed one an d two particle densities , in Math.Frosch.Oberwolfach 43 (2004), Large Scale Stochastic Dynamics, ed. C. Landim, S. O lla, and H. Spohn; O. Costin and J. L. Lebowitz, On the construction of particle distributions with speciﬁe d single and pair densities , J....

View on Pith 2004

[19] [19]

G. Parisi and F. Zamponi, The ideal glass transition of hard spheres , J. Chem. Phys. 123, 144501 (2005)

View on Pith 2005