{"id":"f9aa16c4-b818-4af6-ade0-f8d9c3822c87","arxiv_id":"2412.15070","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A double-holographic Page curve is realized as an entanglement membrane with a vertical growing segment before the Page time and two saturated butterfly-velocity lines exiting through a boundary after it.","lead":"This paper shows that the Page curve of Hawking radiation in a double-holographic black hole model is reproduced by the entanglement membrane, a coarse-grained description of entanglement growth in chaotic quantum systems. It builds a quantitative dictionary between a semiclassical gravity calculation and a many-body calculation of black hole information loss.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The generic-θ extension in §4.2 is unquantified: without controlling x_b and O(1) backreaction corrections, the central claim holds only in the tensionless probe-brane limit.","rationale":"The reader's conditionality is justified. The probe-brane calculation is a coherent application of the holographic membrane construction of [54, 56]: the HM surface projects to a v = 0 membrane and the static island surfaces to ±v_B lines exiting at the orbifold plane, giving (4.4) with t_P = b/v_E. Within that limit the argument is standard. The load-bearing extension is Section 4.2: the claim that for θ < π/2 the same membrane theory results to O(Λ) requires the brane backreaction to affect only O(1) subleading data. The paper has no quantitative control of this: x_b is not computed, and footnote 20 explicitly rests on belief. Since the abstract is not restricted to the T = 0 probe brane, a failure of this step would reduce the result to a special-case illustration. The numerical metric of [33] is available, so the assumption is testable. I also noticed a self-contained typo in the displayed v_E formula in (4.1); solving (2.5) gives v_E = sqrt(d/(d-2)) [(d-2)/(2(d-1))]^{(d-1)/d}, and the printed expression violates v_E ≤ v_B for d = 3. This is secondary and should be corrected, but it does not change the verdict.","tokens_in":18661,"tokens_out":11587,"duration_ms":100527,"concrete_test":"Use the numerical AdS/BCFT black-hole metrics of [33] (d = 4, for a few θ < π/2 such as θ = π/4 and π/6) to compute the island HRT surfaces for reservoir offsets b = 10, 20, 40 in horizon units. Project each surface to the boundary along constant infalling time u and measure (i) the projected exit coordinate x_b, (ii) the slope of the membrane in the plateau region, and (iii) the area S. If S differs from 2 s_th vol(∂A) b by an amount that grows linearly in b, or if the projected slope is not v_B to O(1/b), then the θ < π/2 universality claim fails. If the corrections are O(1) in b and x_b is O(1), the conditional acceptance stands. As a separate sanity check, recompute the displayed v_E formula in (4.1) from (2.5); the printed expression appears to violate v_E ≤ v_B for d = 3.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weakest point is the step from the exactly solvable probe-brane limit (θ = π/2, Section 4.1) to the claimed generic-θ statement (Section 4.2). The abstract's central claim is not restricted to T = 0: it asserts a quantitative equivalence for a double-holographic model in d > 2. To extend the membrane result, the paper needs the backreacted brane geometry to change the island HRT surfaces only at O(1) when b >> Γ. But the paper does not compute x_b, the projected endpoint of the brane, and explicitly relies on belief in footnote 20 ('We believe that ... the difference among this family of extremal surfaces is an O(1) effect'). Since no analytic metric for θ < π/2 is known, only the numerical d = 4 solution of [33], the statement that the majority of island surfaces lie away from the brane and receive subleading corrections is an assertion rather than a proved limit. If x_b or the area corrections were O(Λ) rather than O(1), the membrane equivalence would be restricted to the tensionless probe brane, and the unqualified d > 2 claim in the abstract would fail. The probe-brane derivation itself is coherent, but it does not by itself establish the generic-θ claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a derivation of entanglement membrane theory from a double-holographic model of an eternal two-sided asymptotically AdS_d (d>2) black hole coupled to flat d-dimensional baths. In the probe-brane limit θ=π/2, where the bulk is exactly planar AdS_{d+1}-Schwarzschild, the authors show that the Hartman-Maldacena surface projects to a v=0 membrane and the island surfaces project to ±v_B membranes exiting through the x=0 boundary, yielding the Page curve (4.4) with Page time t_P=b/v_E. For generic θ<π/2 they argue the same membrane description holds to O(Λ) when b≫Γ, where Γ is the backreaction length scale, because the brane backreaction and the unknown extension boundary x_b are O(1) effects. They compare the resulting membrane picture to the Blake-Thompson random-circuit model and obtain a matching dictionary. An appendix applies joining-quench membrane theory to evaporating black holes before the Page time.","tokens_in":19015,"tokens_out":13533,"duration_ms":119350,"significance":"If the central derivation is accepted, the paper provides a concrete and economical bridge between semiclassical island calculations and the modern entanglement-membrane framework: the membrane parameters v_E and v_B are not fitted but inherited from holographic membrane theory, and the saturated membrane configuration realizes the finite-system 'exit through boundary' mechanism. The probe-brane calculation is clean and largely analytic, and the paper is honest about the heuristic nature of the generic-θ extension. The main value is as a proof of principle that double holography can be projected into membrane language, and as a useful dictionary for future comparisons. However, the abstract's unqualified d>2 claim is stronger than what is actually established.","major_comments":[{"comment":"The extension from the exactly solvable θ=π/2 probe-brane limit to generic θ<π/2 is load-bearing for the abstract and the concluding claim, but it is not established. The text asserts that for b≫Γ the backreaction of the Planck brane and the unknown extended boundary x_b are O(1) effects, so that the island surfaces and the HM surface are unchanged to O(Λ). However, x_b is never computed, Γ is not defined quantitatively, and footnote 20 explicitly relies on belief rather than proof for the O(1) nature of the differences within the continuous family of extremal surfaces. Because no analytic metric for θ<π/2 is known (only the d=4 numerical solution of [33]), the statement that the membrane theory and the Page curve (4.4) hold for generic θ is an unproven assertion. This should be either proved/quantified using the numerical metric or removed from the abstract and conclusion by restricting the central claim to the tensionless probe-brane case.","section":"§4.2, footnote 20"},{"comment":"The same missing control affects the claim that before the Page time the HM surface is unaffected by the brane backreaction to O(Λ). The sentence 'the HM surface will be located in regions where the (d+1)-dimensional bulk geometry is well approximated by the AdS_{d+1}-Schwarzschild black hole' is plausible but unquantified; no argument shows that the portion of the surface near the brane contributes at most O(1) to the area or that the projected membrane endpoint is not shifted by an O(Λ) amount. Since the membrane projection is itself only valid to O(Λ) in the scaling limit (2.4), an O(1) ambiguity in the brane region is not obviously harmless. A quantitative estimate from the numerical solution of [33], or a restriction of the claim to the probe-brane limit, is needed.","section":"§4.2, HM surface paragraph"}],"minor_comments":[{"comment":"The relation between the reservoir distance b and the QES location a appears to have a wrong coefficient. From (2.12) and the definition v_B = sqrt(-f'(1)/(2(d-1))), one obtains b ~ -log(1-a)/(2(d-1)v_B), hence a ≈ 1 - exp[-2(d-1)v_B b], not 1 - exp[-d v_B b] as written. The two expressions differ for all d>2; please check and correct.","section":"Eq. (4.6)"},{"comment":"The length scale Γ is introduced only as 'the distance scale over which the backreaction of the brane has a significant effect on the geometry'. Since the condition b≫Γ is essential to the argument, Γ should be defined in terms of the brane angle θ, the horizon radius, and the black-hole parameters, or at least estimated for the numerical solution of [33].","section":"§4.2"},{"comment":"There are a number of typos and minor wording issues: 'As such the the portion of island surfaces' in §4.2; 'generalisating' in §1; 'forevaporating' in §5; 'fine-graned' in §5. These should be corrected in a final pass.","section":"Throughout"},{"comment":"The word 'equivalence' in the abstract may overstate what is shown, because the membrane tension and velocities are inherited from the holographic membrane theory and the match to Blake-Thompson is made by identifying S_BH = s_th b after the fact. I suggest saying 'correspondence' or explicitly describing the dictionary.","section":"§4.3, abstract"}],"recommendation":"major_revision","confidential_remarks":"The probe-brane derivation is sound and the paper is honest about the heuristic nature of the generic-θ extension. The main risk is that the abstract overclaims: if the O(1) statements in §4.2 cannot be made quantitative, the paper should be reframed as a derivation in the tensionless probe-brane limit plus a conjecture for θ<π/2. The Eq. (4.6) coefficient error should also be fixed. With those changes I would expect the paper to be publishable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this paper gives a clean, analytic derivation of the Page curve in a double-holographic model as an entanglement membrane calculation, in the probe-brane limit (θ = π/2). The HM surface projects to a v = 0 membrane, the island surfaces to ±v_B lines exiting at the boundary, and the result is S = 2 s_th vol(∂A) min(v_E t, b). That is a genuinely new dictionary between semiclassical island computations and the Blake–Thompson membrane model.\n\nWhat is good: the probe-brane calculation is coherent and self-contained. It uses standard membrane technology from Mezei and others, applies it to the double-holography setup, and the geometric picture (island surfaces as half-RT surfaces orbifolded at x = 0) is worked out carefully. The comparison to Blake–Thompson is made quantitative by identifying S_BH = s_th b, which is a parameter dictionary rather than a prediction, but it is clearly presented as such. The authors are also honest about what is not proven: Section 4.2 is explicitly an argument, Appendix A is labeled heuristic, and footnote 20 says 'we believe' rather than 'we prove.' Citations are appropriate, with the key prior work ([33], [35], [47], [54]) correctly used.\n\nThe soft spot, and it is the only serious one, is the extension to generic θ < π/2. The argument requires b ≫ Γ so that the brane backreaction is a local effect, and then asserts that the island surfaces are affected only at O(1). But the paper does not compute x_b, the endpoint of the extended boundary, and the area corrections are not quantified. If those corrections turn out to be O(Λ) rather than O(1), the unqualified d > 2 claim in the abstract would be restricted to the tensionless probe brane. This does not sink the probe-brane result: the generic-θ claim is clearly flagged as an argument, and the probe-brane derivation stands on its own. But the abstract overreaches relative to what is proven.\n\nFor a referee: I would accept this for review. The core result is solid and worth publishing; the main request would be to qualify the abstract and either weaken Section 4.2 to a conjecture or provide an error estimate using the numerical solution of [33]. The evaporating black hole appendix is fine as an outlook.\n\nWho this is for: anyone working on islands, the entanglement membrane, or the black hole information problem in d > 2. It deserves a serious referee and, after the θ < π/2 claim is tempered, publication in JHEP or similar. I would cite it for the probe-brane membrane dictionary.","headline":"Probe-brane membrane derivation of the Page curve is clean and new; the generic-θ extension is an openly flagged scaling argument that needs a quantitative check.","tokens_in":19444,"tokens_out":4690,"would_cite":true,"duration_ms":34228,"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":"The paper establishes that, in a double-holographic model of an eternal black hole coupled to baths, the late-time Page curve is quantitatively the entanglement-membrane Page curve of a chaotic many-body system.","keywords":["entanglement membrane","double holography","Page curve","quantum extremal islands","Hawking radiation","AdS/BCFT","Planck brane","entanglement entropy"],"falsifier":"Compute the exact island HRT surface in the backreacted $\\theta<\\pi/2$ geometry (for instance from the numerical $d=4$ solution) at $b\\gg\\Gamma$, project it along constant infalling time, and compare the projection with the $\\pm v_B$ lines and the area with $2s_{\\rm th}\\mathrm{vol}(\\partial A)b$ at leading order in the large-$b$ scaling. Any order-$\\Lambda$ difference would restrict the membrane equivalence to the tensionless probe brane.","tokens_in":18452,"feed_emoji":"🕳️","tokens_out":10301,"duration_ms":84720,"temperature":0.7,"pith_summary":"This paper claims that a semi-classical gravity computation of the Page curve and a coarse-grained chaotic many-body computation are the same calculation, not merely analogous ones. The arena is a double-holographic model: an eternal two-sided AdS$_d$ black hole with $d>2$, each side coupled to a flat non-gravitating bath, so that the quantum extremal island rule is geometrized by ordinary extremal surfaces in one higher dimension. In the late-time, large-subregion limit those extremal surfaces project to entanglement membranes and give $S(R)=2s_{\\rm th}\\mathrm{vol}(\\partial A)\\min(v_E t,b)$, with Page time $t_P=b/v_E$. The paper then shows that this is the Page curve of the Blake-Thompson entanglement-membrane model, thereby making the gravity-vs-quantum-chaos correspondence quantitative.","feed_headline":"Gravity and chaos give the same Page curve","feed_subtitle":"The island-rule calculation and the entanglement-membrane calculation now match, with Page time b/v_E.","key_machinery":"The central object is the entanglement membrane, a coarse-grained surface whose minimal action computes the time-dependent entanglement entropy. The load-bearing mechanism is double holography followed by projection: the quantum extremal island rule in $d$ dimensions is converted into ordinary HRT surfaces in an AdS$_{d+1}$ bulk, and those surfaces are projected along constant infalling time onto the boundary, which in the large-subregion scaling limit yields the membrane action with tension $E(v)$ fixed by the black-brane geometry. The identity that carries the argument is the Page curve $S(R)=2s_{\\rm th}\\mathrm{vol}(\\partial A)\\min(v_E t,b)$, realized by the competition between a $v=0$ membrane (entanglement growth at speed $v_E$) and two $\\pm v_B$ membranes exiting through the boundary (saturation at the coarse-grained entropy).","core_discovery":"The central claim is that the Page curve in a double-holographic black hole model is exactly the curve of membrane theory. The Hartman-Maldacena surface, which grows linearly in time through the black hole interior, projects along constant infalling time to a $v=0$ membrane; the static island surfaces that end on the Planck brane project to two membranes of slope $\\pm v_B$ that exit through the $x=0$ boundary. Minimizing between these two configurations gives $S(R)=2s_{\\rm th}\\mathrm{vol}(\\partial A)\\min(v_E t,b)$, with $t_P=b/v_E$. This is the same Page curve that was obtained for two chaotic systems each coupled to a bath in the Blake-Thompson model, and the paper gives the dictionary between the two models: equal entropy densities and a reservoir backed off by a distance $b$ from the boundary. For the tensionless probe brane the geometry is exact; for general $\\theta<\\pi/2$ the paper argues that brane backreaction shifts the extremal surfaces only at subleading order when $b$ is large, so the membrane description survives.","pith_inferences":["A direct testable extension: if the $\\theta<\\pi/2$ claim holds, the membrane tension $E(v)$ in this whole family of models is determined by the far-from-brane AdS$_{d+1}$-Schwarzschild region alone; computing $E(v)$ from the exact backreacted metric should show no order-$\\Lambda$ dependence on the brane angle.","The dictionary with the Blake-Thompson model suggests a general rule for double holography: endpoints of HRT surfaces on a codimension-one brane appear as membrane endpoints on an effective boundary, so the brane tension and profile control only subleading data such as the endpoint location.","The joining-quench result in the appendix indicates a concrete next step: once the brane profile for an evaporating black hole is known, the same projection should yield the full evaporating Page curve in $d>2$, extending the equivalence beyond eternal black holes."],"forward_implications":["The Page time in this setup is exactly $t_P=b/v_E$: for a given reservoir size $b$, the crossover from linear growth to saturation is controlled by the entanglement velocity of the dual field theory.","The saturated entropy is $2s_{\\rm th}\\mathrm{vol}(\\partial A)b$, which is the coarse-grained entropy of the gravitating system including the large piece of bath; this makes the saturation value a geometric quantity.","The Planck brane acts as a boundary through which saturated membranes can exit, giving a concrete realization of the membrane-theory saturation mechanism in a gravitational setting.","In $d>2$ the ordinary, non-degenerate membrane theory applies, so the double-holographic Page curve is captured by the standard membrane action rather than the generalized 2d CFT version."],"supporting_citations":[{"why":"It supplies the double-holographic mechanism that turns the quantum extremal island rule into ordinary HRT surfaces in one higher dimension.","marker":"[22]"},{"why":"It provides the d>2 double-holographic black-hole-plus-bath setup and the numerical backreacted brane geometry used for the general brane-angle case.","marker":"[33]"},{"why":"It defines the tensionless probe-brane limit in which the bulk is exactly planar AdS$_{d+1}$-Schwarzschild and the island surfaces become tractable.","marker":"[35]"},{"why":"It supplies the chaotic many-body entanglement-membrane model whose Page curve the gravity calculation is claimed to reproduce.","marker":"[47]"},{"why":"It introduces the coarse-grained membrane picture and the saturation mechanism in which membranes of slope $\\pm v_B$ exit through system boundaries.","marker":"[48]"},{"why":"It derives membrane theory from holography by projecting HRT surfaces along constant infalling time, producing the tension $E(v)$ and the velocities $v_E$ and $v_B$.","marker":"[54]"},{"why":"It shows that near-horizon static RT surfaces project to lines of slope $\\pm v_B$, which identifies the island surfaces with exiting saturated membranes.","marker":"[56]"},{"why":"It gives the Hartman-Maldacena interior surface whose linear growth at the entanglement velocity becomes the $v=0$ membrane.","marker":"[34]"}],"fun_headline_variants":["Island rule equals entanglement membrane Page curve","Gravity's island rule matches chaos membrane theory","Page curve identical from holography and chaos","Double holography unifies gravity and chaos","Same Page curve: island rule and membrane model"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"For $\\theta<\\pi/2$, the argument assumes that when the reservoir is very large compared with the backreaction scale of the brane, the brane changes the island surfaces only by order-one amounts, so their leading-order projections and areas match the tensionless probe-brane case; the paper does not compute the extended boundary location and relies on this expectation rather than a proof.","fun_headline_variants_meta":{"raw":{"variants":["Island rule equals entanglement membrane Page curve","Gravity's island rule matches chaos membrane theory","Page curve identical from holography and chaos","Double holography unifies gravity and chaos","Same Page curve: island rule and membrane model"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000554,"raw_usage":{"total_tokens":2610,"prompt_tokens":888,"completion_tokens":1722,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":504,"completion_tokens_details":{"reasoning_tokens":1652}},"tokens_in":504,"tokens_out":1722,"duration_ms":11389,"temperature":1.0,"reasoning_tokens":1652,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:40:04.160384+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the exact island HRT surface in the backreacted $\\theta<\\pi/2$ geometry (for instance from the numerical $d=4$ solution) at $b\\gg\\Gamma$, project it along constant infalling time, and compare the projection with the $\\pm v_B$ lines and the area with $2s_{\\rm th}\\mathrm{vol}(\\partial A)b$ at leading order in the large-$b$ scaling. Any order-$\\Lambda$ difference would restrict the membrane equivalence to the tensionless probe brane.","supporting_citations":[{"cited_title":"The Page curve from the entanglement membrane","cited_arxiv_id":"2306.13140","evidence_quote":"It supplies the chaotic many-body entanglement-membrane model whose Page curve the gravity calculation is claimed to reproduce."},{"cited_title":"The entanglement membrane in 2d CFT: reflected entropy, RG flow, and information velocity","cited_arxiv_id":"2411.16542","evidence_quote":"It shows that near-horizon static RT surfaces project to lines of slope $\\pm v_B$, which identifies the island surfaces with exiting saturated membranes."}],"review_version":1}