{"id":"7e41bb0a-2c42-4800-a28a-d18c75e26e64","arxiv_id":"2507.22120","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A family of time-symmetric Kaluza-Klein bubble initial data exists with arbitrarily negative ADM energy for fixed bubble and circle sizes.","lead":"This paper constructs explicit five-dimensional gravity solutions, small Kaluza-Klein bubbles of nothing, whose total ADM energy can be made arbitrarily negative while the asymptotic circle radius and the bubble size stay fixed. The result means the Kaluza-Klein vacuum may be more unstable than previously thought, and it highlights a sharp contrast with the bounded energy required by AdS/CFT.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The p^3 energy formula stands or falls on the unproven I2=O(p) estimate in Eq. (13c); the heuristic and Fig. 1 do not exclude a long near-zero plateau that would make I2~-p^3 and cancel the leading negative term.","rationale":"Agree with the reader's weakest assumption: the I2 estimate is the load-bearing step. I found the rest of the argument coherent: the choice of alpha, the smoothness condition, the large-p expansion of e^{I1} in Eqs. (23)-(25), and the Brill-Pfister consistency check are explicit and mutually consistent. The I2 paragraph is the only place where the paper passes from exact formulae to a magnitude claim supported by a heuristic and one numerical figure. I do not claim the estimate is wrong; the asymptotic forms (A6)-(A7) make it plausible, and the sign structure means only a long near-zero plateau of the positive integrand, not an overshoot, could destroy the result. But because no bound is supplied, the central claim is not yet fully established. This does not warrant rejection: the construction is explicit, the numerics are reasonable, and the risk is medium, as the reader judged. It does warrant a conditional posture until the I2 scaling is checked independently or proved; if the check confirms I2=O(p), the ACCEPT verdict is fully justified.","tokens_in":11588,"tokens_out":12309,"duration_ms":159538,"concrete_test":"Use arbitrary-precision integration in u=1/rho, with the tail subtracted using Eq. (A7), to compute I2 for p=10^2,10^3,10^4 at fixed rho0=d=R=1. Also record the crossing scale y*(p)=min{y>=rho0: F(y)=1/2} for F the positive first factor in the I2 integrand. If |I2|<=C p and y*(p)/p is bounded for all three p, the heuristic is confirmed and Eq. (27) stands. If instead |I2| scales like p^3 with a negative sign, so that -I2/2 gives a positive p^3 contribution with coefficient at least 1/(2^8 R^2), the claimed unboundedness would fail and the paper should be revised.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is Eq. (27): E = -p^3/(2^9 R^2)+O(p^2). The derivation up to Eq. (26) is explicit; the only step that is not a controlled expansion is the estimate of I2 in Eq. (13c), namely that I2=O(p). The paper's justification is the sentence before Fig. 1: the positive first factor of the I2 integrand stays close to zero until rho ~ sqrt(|alpha2|) ~ p and then approaches 1, so the integrand contributes at most a length-p region. This is a heuristic, not a bound: no pointwise estimate is proved for the factor F(y)=rho0 exp(int_{rho0}^y h1/h2)/(e^{I1}h2(y)), and Fig. 1 only goes to p=100 and, as the paper's own footnote states, is a difference of divergent terms with a cutoff. A positive overshoot of F would only make I2 positive and thus make E more negative, so the dangerous scenario is a long interval, of length ~p^3, on which F is near zero; then I2 ~ -c p^3 and -I2/2 contributes +c p^3/2, which cancels Eq. (26) when c >= 1/(2^8 R^2). Nothing in Section IV rules out that scenario rigorously. This is not a criticism of the construction itself, but it is the place where the central claim is least secure.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper constructs an explicit family of time-symmetric, SO(3) x U(1)-symmetric initial data for five-dimensional Kaluza-Klein theory and analyzes the ADM energy of the resulting bubble-of-nothing spacetimes. The metric ansatz is given in Eq. (2), with the function alpha(rho) chosen in Eq. (14). The authors solve the Hamiltonian constraint exactly through Eq. (10) and derive a general closed-form expression for the ADM energy, Eq. (13). For large values of the parameter p, they obtain the asymptotic result E = -p^3/(2^9 R^2) + O(p^2), with the bubble radius rho0 and the S1 radius R fixed, implying that the ADM energy is unbounded from below even for fixed geometric bubble size. The paper also checks consistency with the Brill-Pfister bound, shows that the bubbles initially expand, and presents numerical evidence that the analogous construction in asymptotically AdS spacetimes does not lead to arbitrarily negative energy.","tokens_in":11935,"tokens_out":6573,"duration_ms":77321,"significance":"If the central estimate is correct, the paper settles a natural question in Kaluza-Klein gravity: it shows that no lower bound on the ADM energy can be expressed solely in terms of the size of the minimal S2 at the bubble and the S1 radius at infinity. The construction is explicit, the Hamiltonian constraint is solved exactly, and the leading term in Eq. (27) is derived analytically. The paper also gives useful consistency checks (Section V) and a physically motivated contrast with AdS/CFT (Section VI), where the same construction yields bounded energy. A particular strength is that the parameter p is a free parameter of the ansatz, not fitted to the target energy, so the argument is not circular. The main weakness is that the estimate I2 = O(p) in Section IV, on which the unboundedness result depends, is supported only by a heuristic argument and a numerical integration with a cutoff rather than by a rigorous bound.","major_comments":[{"comment":"The central claim E = -p^3/(2^9 R^2) + O(p^2) relies on the estimate I2 = O(p) for the integral defined in Eq. (13c). The justification in the text before Fig. 1 is heuristic: it asserts that the positive first factor in the integrand stays close to zero until rho ~ sqrt(|alpha2|) ~ p and then approaches one. This does not rule out an interval of length ~p^3 on which that factor is still near zero, which would make I2 ~ -c p^3 and therefore -I2/2 ~ +c p^3/2, potentially canceling the leading negative term in Eq. (26). The numerical plot in Fig. 1 is not a substitute for a proof: it covers only p <= 100 and, as the footnote states, involves a cutoff for divergent terms. I ask the authors to replace this step with a rigorous pointwise bound on the integrand, using the explicit expression in Eq. (20), or to provide a controlled asymptotic expansion of I2 as p -> infinity that establishes I2 = O(p).","section":"Sec. IV, Eq. (13c) and Eq. (27)"},{"comment":"The same uncontrolled estimate is used in Appendix B, where the claim that \"by the same reasoning as in section IV, I2 is at most linear in b\" is asserted. The integrand there involves the roots of the quintic Q(x), so the behavior is more complicated than in the main text. Since the unboundedness result in Appendix B also hinges on this linear-in-b estimate, the authors should either prove it or state explicitly that the appendix reports a conjectural family pending a rigorous bound.","section":"Appendix B, Eq. (B16)"}],"minor_comments":[{"comment":"The ansatz states p in R, but the positivity of h2 and the large-p analysis require p >= 0; please state this restriction explicitly.","section":"Eq. (14)"},{"comment":"The roots rho_i of the cubic can be complex, but the notation treats them as real quantities. It would be clearer to state that the partial-fraction decomposition and the products are understood in the complex plane and that the final expressions are real.","section":"Eqs. (17)-(19)"},{"comment":"The figure would be more informative if it showed the numerical error associated with the cutoff and an overlay of a linear fit, so the reader can assess how convincingly the stated I2 = O(p) behavior is supported.","section":"Fig. 1"},{"comment":"The AdS analysis is numerical for selected parameter values; the text appropriately hedges the conclusion, but it would help to list the range of rho0/l and p values that were checked.","section":"Sec. VI"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is well within the scope of the journal and the construction is interesting. The main technical gap is the I2 = O(p) estimate in Section IV and its analog in Appendix B. If the authors can supply a rigorous bound, I would support acceptance. I do not see any concerns about attribution or citation practice."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Horowitz and Lu do something new: they write down an explicit one-parameter family of time-symmetric KK bubble initial data where the ADM energy goes like -p^3 even though both the minimal S2 radius and the S1 radius at infinity are held fixed. Previous Brill-Horowitz examples needed the bubble to grow to make the energy more negative, and Witten's bubble sat at zero energy. If the calculation is right, there is no bound in terms of these geometric sizes, which strengthens the case that the KK vacuum is severely unstable. The construction is concrete and the energy formula (13) is derived in closed form, with the asymptotics pushed through analytically.\n\nThe one place I would want more support is the estimate I2=O(p) that underlies Eq. (27). It is not a fully rigorous bound; the paper gives a qualitative argument about the integrand being exponentially suppressed until rho ~ p and then approaching 1, and backs it with a numerical plot up to p=100. The stress-test note worried about a long near-zero plateau of length p^3 canceling the p^3 term. Having looked at the explicit expression (20), I don't think that scenario is real: the ratio F is 1 - alpha2/(2rho^2)+... asymptotically and the exponential factor is e^{-2p/rho}, so any region where F is far from 1 has length O(p). The estimate is very likely correct, but it is the load-bearing step and a serious referee should ask for a bound or a more detailed asymptotic control. The authors themselves flag it as heuristic, which is honest. The AdS/CFT section is numerical and not presented as a proof; it is a consistency check, not part of the main claim.\n\nI checked the consistency checks: the Brill-Pfister bound is satisfied (Fig. 2), and Appendix B gives a second family where curvature grows only polynomially, which reinforces that the unboundedness is not an artifact of the specific ansatz. Citation behavior is unobjectionable; the background work is properly credited.\n\nBottom line: this is a solid, explicit construction answering a natural question, and it deserves a real referee. The result is likely correct; the I2 estimate should be tightened during revision. If it holds up, it is a cite-worthy result for anyone working on gravitational energy bounds or KK stability.","headline":"A genuinely new unbounded-below energy result for fixed-size KK bubbles; the only unproven step, I2=O(p), looks right on inspection but should be tightened by a referee.","tokens_in":12441,"tokens_out":6631,"would_cite":true,"duration_ms":62603,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A family of time-symmetric initial data makes Kaluza-Klein bubble energy unbounded below even at fixed geometric sizes.","keywords":["Kaluza-Klein bubble of nothing","ADM energy","negative energy","time-symmetric initial data","Hamiltonian constraint","Kaluza-Klein vacuum instability","AdS/CFT","unbounded energy"],"falsifier":"Evaluate $I_2(p)$ from Eq. (13c) by high-precision numerical integration for, say, $\\rho_0=d=1$ and $R=1$ at large $p$ (e.g. $p=10^2$ to $10^4$). If $\\log I_2/\\log p$ approaches $3$ rather than $1$, the claimed $O(p)$ behavior fails and the central conclusion is false.","tokens_in":11370,"feed_emoji":"🫧","tokens_out":8513,"duration_ms":91662,"temperature":0.7,"pith_summary":"The paper tries to establish that the total (ADM) gravitational energy of a Kaluza-Klein bubble of nothing is unbounded from below, even when the two geometric scales one would naturally fix are held fixed: the radius $R$ of the $S^1$ at infinity and the radius $\\rho_0$ of the minimal $S^2$ where the circle pinches off. It does this with an explicit family of time-symmetric initial data controlled by a parameter $p$; for large $p$ the energy is $E = -p^3/(2^9 R^2) + O(p^2)$, so it tends to $-\\infty$ while $\\rho_0$ and $R$ do not change. Because this occurs for arbitrarily small bubbles, the paper concludes that the standard Kaluza-Klein vacuum is more unstable than earlier negative-energy examples suggested. The analogous construction with a negative cosmological constant does not become unbounded, which the authors read as agreement with AdS/CFT.","feed_headline":"Small Kaluza-Klein bubbles can carry arbitrarily negative energy","feed_subtitle":"With the circle and bubble sizes fixed, explicit initial data send the ADM energy to minus infinity.","key_machinery":"The machinery is the ansatz for $\\alpha(\\rho)$ together with the integral solution of the Hamiltonian constraint. With $f_1=(\\alpha\\beta)^{-1}$ and $f_2=\\alpha$, the constraint ${}^{(4)}R=0$ becomes a first-order equation for $\\beta$, solved by $\\beta(\\rho)=e^{-\\int h_1/h_2}(c+\\int e^{\\int h_1/h_2}/h_2)$. The chosen $\\alpha$ makes the large-$p$ expansion of $I_1$ produce a term $e^{I_1} \\sim 2^{10}\\rho_0 d^2 p^{-3} e^{2p/\\rho_0}$, which, when multiplied by $(\\alpha'(\\rho_0))^2=e^{-2p/\\rho_0}/d^2$, yields the cubic negative term. The argument that the positive integral $I_2$ does not cancel it rests on the asymptotic behavior of the integrand and on numerical integration.","core_discovery":"The central claim is that no bound on the ADM energy of a Kaluza-Klein bubble of nothing can be formed from the bubble's geometric size. Starting from a time-symmetric metric $ds^2 = (\\alpha\\beta)^{-1} d\\rho^2 + \\rho^2 d\\Omega^2 + \\alpha d\\phi^2$ and imposing the Hamiltonian constraint, the paper chooses $\\alpha(\\rho) = (\\rho-\\rho_0)/(\\rho-\\rho_0+d)\\,e^{-p/\\rho}$. The no-conical-singularity condition fixes the integration constant $c = 4(\\alpha'(\\rho_0))^2/R^2$, and the large-$p$ asymptotics of the integral representation give $e^{I_1} \\simeq 2^{10}\\rho_0 d^2 p^{-3} e^{2p/\\rho_0}$. Since $\\alpha'(\\rho_0)=e^{-p/\\rho_0}/d$, the smoothness term contributes $-p^3/(2^9 R^2)$ to leading order, while all other terms contribute at most $O(p^2)$. The result is Eq. (27), $E = -p^3/(2^9 R^2) + O(p^2)$, with $\\rho_0$, $R$, and $d$ fixed.","pith_inferences":["A rigorous bound showing $I_2=O(p)$ would turn the heuristic cancellation argument into a proof; until then the unboundedness claim rests on that numerical and heuristic step.","Because the main family's curvature at the bubble grows exponentially with $|E|^{1/3}$, quantum gravity may restore a lower bound at the Planck scale; the appendix's polynomial-curvature family indicates that this is not an artifact of the specific ansatz.","The AdS contrast suggests a general energy-minimization theorem for fixed boundary $S^2\\times S^1\\times R$; extending recent proofs from other boundary topologies could settle it.","Numerically evolving this initial data would test whether the combination of a small bubble and huge negative energy leads to rapid expansion or to a naked singularity."],"forward_implications":["No lower bound on the ADM energy of a Kaluza-Klein bubble of nothing can be stated in terms of the circle radius at infinity and the minimal-sphere radius at the bubble.","Arbitrarily small bubbles can have arbitrarily negative energy, so the flat Kaluza-Klein vacuum is unstable in a stronger sense than the older negative-energy examples suggested.","Zero-energy initial data that look like the Kaluza-Klein vacuum outside a small region should exist, making a small-action instanton for vacuum decay plausible.","The AdS analogue remains bounded below, so the mechanism is special to asymptotically flat Kaluza-Klein boundary conditions, consistent with AdS/CFT.","Bubbles with very negative energy start out expanding at the moment of time symmetry, since very negative energy forces $\\alpha'(\\rho_0)$ to be small."],"supporting_citations":[{"why":"It constructs the original bubble-of-nothing solution with zero energy, the object whose negative-energy deformations are studied here.","marker":"[10]"},{"why":"It produces negative-energy Kaluza-Klein bubbles whose energy is bounded as the bubble grows, the family this paper extends past that bound.","marker":"[11]"},{"why":"It establishes the Brill-Pfister lower bound in a conformal radial coordinate, the bound the new family must and does satisfy consistently.","marker":"[12]"},{"why":"It gives the ADM energy formula for five-dimensional gravity with compactified dimensions used in Eq. (1).","marker":"[13]"},{"why":"It provides the gravitational energy expression in spacetimes with compactified dimensions that supports the energy formula.","marker":"[14]"},{"why":"It gives the initial acceleration formula for a bubble at time symmetry, used to show the new bubbles start out expanding.","marker":"[15]"},{"why":"It finds the static AdS solutions with $S^2\\times S^1\\times R$ boundary and conjectures the lowest-energy one minimizes the energy, the target of the AdS contrast.","marker":"[16]"},{"why":"It derives a refined energy inequality in AdS spacetimes that the flat-space unboundedness must not violate.","marker":"[17]"}],"fun_headline_variants":["Bubble energy unbounded below","Small Kaluza-Klein bubbles: energy to minus infinity","No floor on bubble of nothing energy","KK bubbles destabilize the vacuum"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the integral $I_2$ in the energy formula grows at most linearly in the parameter $p$; if it grew as fast as $p^3$, it could cancel the negative leading term and the energy would not be unbounded.","fun_headline_variants_meta":{"raw":{"variants":["Bubble energy unbounded below","Small Kaluza-Klein bubbles: energy to minus infinity","No floor on bubble of nothing energy","KK bubbles destabilize the vacuum"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1653,"prompt_tokens":890,"completion_tokens":763,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":506,"completion_tokens_details":{"reasoning_tokens":709}},"tokens_in":506,"tokens_out":763,"duration_ms":9160,"temperature":1.0,"reasoning_tokens":709,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T12:03:29.784923+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate $I_2(p)$ from Eq. (13c) by high-precision numerical integration for, say, $\\rho_0=d=1$ and $R=1$ at large $p$ (e.g. $p=10^2$ to $10^4$). If $\\log I_2/\\log p$ approaches $3$ rather than $1$, the claimed $O(p)$ behavior fails and the central conclusion is false.","supporting_citations":[{"cited_title":"Witten, Instability of the Kaluza-Klein vacuum, Nucl","cited_arxiv_id":null,"evidence_quote":"It constructs the original bubble-of-nothing solution with zero energy, the object whose negative-energy deformations are studied here."},{"cited_title":"Brill and G","cited_arxiv_id":null,"evidence_quote":"It produces negative-energy Kaluza-Klein bubbles whose energy is bounded as the bubble grows, the family this paper extends past that bound."},{"cited_title":"Arbitrarily Negative Energy for Small Kaluza-Klein Bubbles","cited_arxiv_id":"2507.22120","evidence_quote":"It establishes the Brill-Pfister lower bound in a conformal radial coordinate, the bound the new family must and does satisfy consistently."},{"cited_title":"Dai, A note on positive energy theorem for spaces with asymptotic SUSY compactification, Journal of Mathematical Physics 46, 10.1063/1.1862095 (2005)","cited_arxiv_id":null,"evidence_quote":"It gives the ADM energy formula for five-dimensional gravity with compactified dimensions used in Eq. (1)."},{"cited_title":"Deser and M","cited_arxiv_id":null,"evidence_quote":"It provides the gravitational energy expression in spacetimes with compactified dimensions that supports the energy formula."},{"cited_title":"We can calculate (4)Rθθ(ρ0) = 1 − ρ0α′(ρ0)β(ρ0) = 1 − 4ρ0 R2α′(ρ0) , (34) where we have used Eq","cited_arxiv_id":null,"evidence_quote":"It gives the initial acceleration formula for a bubble at time symmetry, used to show the new bubbles start out expanding."},{"cited_title":"Brill and H","cited_arxiv_id":null,"evidence_quote":"It finds the static AdS solutions with $S^2\\times S^1\\times R$ boundary and conjectures the lowest-energy one minimizes the energy, the target of the AdS contrast."},{"cited_title":"Bombelli, R","cited_arxiv_id":null,"evidence_quote":"It derives a refined energy inequality in AdS spacetimes that the flat-space unboundedness must not violate."}],"review_version":1}