{"id":"e7056c83-ce23-4b79-85e0-e3b28230dbf7","arxiv_id":"2412.07549","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"For a Haar-random model of an evaporating black hole, the q-partite Renyi multi-entropy of the black hole plus q-1 radiation subsystems peaks at a multi-entropy time later than the Page time and does not vanish at complete evaporation.","lead":"The paper divides Hawking radiation into several pieces and computes a multipartite entanglement measure, the multi-entropy, as the black hole evaporates, using a random quantum state as the model. It finds that this multi-entropy peaks later than the usual Page time and stays nonzero at the end of evaporation, pointing to hidden entanglement among Hawking particles.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The physically load-bearing step is identifying each point of the curve with a fresh Haar-random state; because the model has no time evolution, the nonzero final multi-entropy is an input of the random-state assumption rather than a derived property of evaporation.","rationale":"The paper is internally consistent and transparent: exact finite (q,n) replica sums are checked, the 1/dBH variance cancellation is demonstrated, and Sections 5.2 and 6 explicitly label the late-time and endpoint formulas as conjectures and list open issues. Those limitations support a conditional verdict rather than a rejection. However, the strongest physical statement—that nonzero final multi-entropy represents secret entanglement between Hawking particles—is not a theorem about evaporation; it is an inference from the random-state model. Since each curve point is computed from an independent Haar-random tensor, the model has no time evolution, and the endpoint value is essentially the concentration-of-measure value for a random pure state on the radiation subsystems. This is exactly the reader's weakest assumption, and it is the most load-bearing part of the argument: if the approximation fails in a real evaporating system, the quantitative curve and the qualitative 'secret entanglement' conclusion do not follow. I would therefore keep the CONDITIONAL verdict, adding a concrete unitary simulation or gravitational derivation as the natural condition for upgrading it.","tokens_in":38708,"tokens_out":9208,"duration_ms":102573,"concrete_test":"Run a minimal unitary evaporation simulation: start from a Haar-random pure state on a black-hole register of dimension D0. At each step, apply a Haar-random isometry V_t: H_BH(t) → H_BH(t+1) ⊗ H_rad from a freshly emitted radiation register, with dim H_BH(t+1) = D0 / dR(t)^(q-1), distributing emitted degrees of freedom equally among q-1 radiation registers. Compute the (q=3, n=2) multi-entropy of the evolved pure state as a function of dR, using Eq. (2.17). If the resulting curve (maximum location and dBH = 1 value) coincides with the static single-random-tensor curve (3.16), the random-state approximation is supported; if the unitary trajectory deviates (e.g., peak shifted or endpoint smaller), the central claim fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central physical claims—that Hawking radiation develops multipartite 'secret entanglement' that survives to complete evaporation—depend on the single-random-tensor model of Section 1 and Eq. (3.4). In that model, each value of dR corresponds to a freshly drawn Haar-random state on H_R1 ⊗ ... ⊗ H_R(q-1) ⊗ H_BH with dR1 = ... = dR(q-1) = dR and dBH = dTotal / dR^(q-1). There is no unitary map connecting the states at different dR, no Hamiltonian, and no emission mechanism. The curve is therefore an ensemble of unrelated equilibrium states, not a time history. At the endpoint dBH = 1 the model simply places a Haar-random pure state on the q-1 radiation subsystems; for q-1 ≥ 2 such a state has O(log dTotal) multi-entropy essentially by concentration of measure. So the reported nonzero endpoint and the interpretation as entanglement between actual Hawking particles are not consequences of black-hole dynamics—they are the assumed random-state model. This is not an internal contradiction; the paper labels the approximation. But it is the load-bearing assumption: if a real evaporating state at fixed dR differs systematically from the Haar-random state (e.g., because of energy conservation, locality, or the pairing structure of Hawking emission), the multi-entropy time and the final value need not match the curves in Figs. 6–12.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the multipartite entanglement structure of Hawking radiation by dividing the radiation into q−1 subsystems and approximating the combined system of radiation plus black hole as a single q-partite Haar-random state. The authors define a 'black hole multi-entropy curve' as the Rényi multi-entropy as a function of the radiation dimension dR, with the black hole dimension fixed by dTotal = dR^{q−1} dBH. They compute exact multi-entropies for (q,n) = (3,2), (3,3), and (2,4), derive the early-time expansion in 1/dBH for general q and n, and conjecture a late-time scaling. The claimed qualitative features are: the multi-entropy increases at early times, peaks at a 'multi-entropy time' dR = dTotal^{1/q} later than the Page time, then decreases but does not vanish at complete evaporation (dBH=1); the nonzero final value is interpreted as 'secret entanglement' between Hawking particles. The paper also compares the multi-entropy curve with Rényi entanglement negativity and reflected entropy.","tokens_in":39027,"tokens_out":18050,"duration_ms":143073,"significance":"If the central claims hold, the paper provides a natural multipartite generalization of the Page curve and suggests that the multipartite entanglement of Hawking radiation can have a different, richer time evolution than the bipartite Page curve predicts. The exact finite-case calculations are carefully performed and include consistency checks, and the early-time variance cancellation is shown explicitly. The paper is commendably honest about its conjectural elements: the late-time scaling is labeled as unproven in Section 5.2, the analytic continuation to n→1 is acknowledged as subtle in Section 2.2, and the random-tensor modeling assumption is stated in the introduction. However, the headline claims are stated for arbitrary n and q and for the von Neumann limit, while the proofs are limited to specific integer cases; the physical interpretation rests on the untested random-state approximation.","major_comments":[{"comment":"The abstract and Section 5.3 present the nonzero final multi-entropy and the multi-entropy time as general features of black hole multi-entropy curves, but the exact calculations are limited to (q,n)=(3,2), (3,3), and (2,4). The late-time behavior for general q and n is a conjecture (Section 5.2, Eqs. (5.21)-(5.22)), and the analytic continuation from integer n to n→1 is explicitly stated to be subtle and not rigorously performed (Section 2.2). The claims in the abstract should be restricted to the Rényi index n for which they are proven, or the proofs should be supplied.","section":"Abstract and Section 5.3"},{"comment":"The 'multi-entropy time' is defined as the point where the multi-entropy is maximal and is asserted to be dR=(dTotal)^{1/q}. For the exactly solved cases this is supported only by numerical plots (Figs. 6-9), not by a derivative analysis of, e.g., Eq. (3.16). For general q,n, the claimed location follows from the intersection of the early-time expansion (5.9) and the conjectured late-time expression (5.21), which does not by itself locate the maximum of the exact function. Since the multi-entropy time is a central concept, this requires a proof or at least a verification for the exact cases.","section":"Section 3.3, item 2, and Section 5.3"},{"comment":"The nonzero final multi-entropy is interpreted in the abstract and Section 6 as 'secret entanglement between Hawking particles.' However, the model has no time evolution: each point on the curve is an independent Haar-random state with different bond dimensions, and the dBH=1 endpoint is a random state on the radiation subsystems. The nonzero final value is therefore an assumption of the model, not a derived property of black hole evaporation. To support the physical claim, the authors should either provide a dynamical realization (e.g., a sequential unitary or random-circuit model) that produces the same curve, or explicitly state in the abstract and conclusion that the result is a property of the random-tensor model and may not describe actual Hawking radiation.","section":"Section 1 and Eq. (3.4)"}],"minor_comments":[{"comment":"Equation (3.25) does not appear to follow from (3.23) after the substitution dR1=dR2=dR and dBH1=dBH2=√dBH; the first parentheses contains an extra term 2dR√dBH and the second parentheses omits the dR^2 dBH term. Please check the algebra.","section":"Section 3.4, Eq. (3.25)"},{"comment":"The formulas (5.23)-(5.25) are presented without repeating the conjecture caveat stated in Section 5.2; since the abstract relies on these formulas, the conjectural status should be carried through to the abstract and conclusion.","section":"Section 5.3"},{"comment":"The notation 'd2_R1' and similar expressions (e.g., in Eqs. (3.23) and (3.25)) is ambiguous; these should be rendered as d_{R1}^2 in the final typed version to avoid confusion.","section":"Throughout"},{"comment":"The figure legend does not clearly distinguish the curve for E4 from that for E; please add a clear legend so that the comparison discussed in Section A.3 is easy to follow.","section":"Figure 14"}],"recommendation":"major_revision","confidential_remarks":"The paper is technically sound in its exact finite-case calculations and is honest about its conjectures. The main issue is the gap between the generality of the physical claims and the proven results. The random-tensor model is a toy model; the editor may wish to consider whether the journal's standards require a stronger connection to actual evaporation dynamics for the physical interpretation to be stated as a result. The citation pattern is appropriate; the self-citations by one author are to technical tools (random tensor networks, reflected entropy) and do not appear excessive."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a careful random-tensor computation of a genuinely new object—the black hole multi-entropy curve. If you care about Page-curve generalizations or multipartite entanglement measures, it is worth your time. The exact finite cases ((q,n)=(3,2), (3,3), (4,2)) are checkable, the 1/dBH variance cancellation is a nice technical result, and the early-time expansion and the Z1 proof are clean. The comparison with negativity and reflected entropy in Appendix A is genuinely useful. The authors also do something rare: they separate what is proven from what is conjectured, and they flag the analytic continuation and late-time scaling as open.\n\nThe soft spots are real, but they are mostly acknowledged in the paper. The model draws a fresh Haar-random state at every dR; there is no unitary evolution, no Hamiltonian, and no emission mechanism. The curve is an ensemble of unrelated equilibrium states, not a time history. Consequently, the nonzero endpoint at dBH=1 is essentially concentration of measure for a Haar-random state on the radiation subsystems—it is a property of the random-state assumption, not a derived consequence of black-hole evaporation. The stress-test note lands on this, and the paper's own Section 1 makes the same caveat. The headline claims are stated for general n and q and include n=1, but the analytic continuation is not proven and the late-time behavior is conjectured in (5.21) and (5.22). So the qualitative inverted-V shape and the multi-entropy time later than Page time are well supported for integer n ≥ 2; the 'secret entanglement' language should be understood as a feature of the model, not a derivation from gravity.\n\nCitation pattern looks fine: self-citations are to technical random-tensor tools, and the discussion of [42] is fair. This is not a paper with hidden fitted parameters or circular logic. It deserves a serious referee. I would send it to review, with the main request that the authors either prove the conjectures or explicitly restrict the general claims to the computed cases, and soften the physical interpretation of the endpoint accordingly.","headline":"A clean, honestly labeled random-tensor calculation of a genuinely new multipartite Page-curve analog; the finite cases are solid, the general-n and n=1 claims are conjectures, and the nonzero endpoint is largely a feature of the Haar-random modeling assumption.","tokens_in":39534,"tokens_out":2114,"would_cite":true,"duration_ms":22758,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.70.Dy","03.67.Mn"],"model":"deepseek-v4-flash","headline":"Multipartite entanglement in Hawking radiation keeps growing past the Page time and never vanishes at full evaporation.","keywords":["black hole information paradox","Page curve","multi-entropy","multipartite entanglement","Hawking radiation","Haar-random states","replica trick","random tensor networks"],"falsifier":"Compute the q=3, n=2 multi-entropy in a concrete unitary evaporation model, such as a random unitary circuit or a spin-chain Hamiltonian that couples a black-hole register to radiation registers with fixed total dimension, and check whether $S_2^{{(3)}}$ peaks at dR=(dTotal)^{1/3} and remains nonzero at dBH=1; a peak at the Page time or a vanishing endpoint would disprove the paper's central claim.","tokens_in":38451,"feed_emoji":"🕳️","tokens_out":7659,"duration_ms":72670,"temperature":0.7,"pith_summary":"This paper asks what happens to quantum entanglement among Hawking radiation itself, not just between radiation and the black hole, as a black hole evaporates. It proposes a black hole multi-entropy curve: for q≥3 equal radiation subsystems plus the black hole, plot the q-partite multi-entropy against the radiation subsystem size. The paper reports that the curve rises, peaks later than the standard Page time, and then falls to a nonzero plateau when the black hole has fully evaporated; the leftover value is read as secret entanglement between Hawking particles that Hawking's semiclassical calculation misses. If correct, the information flow out of a black hole is not captured by the bipartite Page curve alone, and multipartite correlations carry a substantial share of the final-state information.","feed_headline":"Black hole multi-entropy peaks later than Page time","feed_subtitle":"Dividing Hawking radiation into pieces reveals entanglement left after the black hole evaporates.","key_machinery":"The central object is the multi-entropy, a q-partite entanglement measure built by laying $n^{{q−1}}$ copies of the reduced density matrix on a (q−1)-dimensional periodic lattice and contracting legs along each axis; for q=2 it reduces to the usual Rényi entropy. The computation models the evaporating black hole plus radiation as one Haar-random q-partite tensor of bond dimensions dR,...,dR,dBH. The replica trick turns the averaged partition function into a sum over the permutation group S_{$n^{{q−1}}$} with weights $d_i^{{-d(g,g_i)}}$, where d is the Cayley distance; the multi-entropy time dBH=dR is where no subsystem dominates the sum, and the nonzero endpoint survives because at dBH=1 the remaining radiation legs still carry nontrivial contraction structure. The late-time formulas are presented as a conjecture with multiplicative coefficients counting degenerate minimizers, with exact coefficients verified for the small cases by explicit sums.","core_discovery":"The paper's central claim is that for any number q≥3 of parties—one black hole plus q−1 radiation subsystems of equal dimension dR, with total Hilbert-space dimension dTotal fixed—the Rényi multi-entropy $S_n^{{(q)}}$ traces a curve that increases at early times, reaches its maximum at the multi-entropy time dR=(dTotal)^{1/q}, and then decreases, but does not vanish at complete evaporation dBH=1. Since the multi-entropy reduces exactly to the entanglement entropy when q=2, this black hole multi-entropy curve is the direct multipartite generalization of the Page curve. The peak occurs later than the Page time dR=(dTotal)^{1/[2(q−1)]} for all q≥3, and the nonzero endpoint is attributed to entanglement shared among Hawking particles themselves, which is invisible in Hawking's semiclassical approximation. The paper establishes this by exact replica sums for the simplest cases (q=3 with n=2, q=3 with n=3, q=4 with n=2) and by analytic early-time expressions and conjectured late-time expressions valid for general n and q.","pith_inferences":["If a unitary, energy-conserving evaporation model replaces the static Haar random state, the peak location and endpoint may change; computing S_2^{(3)} in a random-unitary or spin-chain model would show whether the nonzero final multi-entropy is a genuine evaporation feature or an artifact of the random-state idealization.","The later peak suggests that experimental probes of multipartite entanglement in analog black-hole settings, such as tripartite witnesses, would see a delayed signal compared with the bipartite Page signal.","The conjectured late-time coefficients count non-crossing permutations, so a combinatorial proof of the b_n^{(q)} and c_n^{(q)} formulas would turn the endpoint prediction into a theorem for all n and q.","Extending the curve to unequal radiation subsystems would likely move the multi-entropy time away from (dTotal)^{1/q}, giving a way to test the equal-partition assumption against more realistic anisotropic emission."],"forward_implications":["For q≥3, higher-partite correlations among Hawking quanta continue to grow after the Page time, so the Page time does not mark the end of information accumulation in the radiation.","At the end of evaporation the multi-entropy is large and positive, implying the final radiation state encodes information in multipartite entanglement rather than only in pairwise black-hole–radiation correlations.","As q grows, the multi-entropy time approaches the evaporation endpoint, and in the q→∞ limit the curve coincides with Hawking's monotone prediction, so the amount of hidden multipartite structure depends on how finely the radiation is partitioned.","The qualitative shape—rise, later peak, nonzero endpoint—appears for every Rényi index n and party number q that the paper computes exactly, and the paper conjectures that it survives the n→1 limit."],"supporting_citations":[{"why":"Supplies the Page curve baseline that the black hole multi-entropy curve generalizes to q≥3.","marker":"[2, 3]"},{"why":"Supplies Hawking's semiclassical radiation prediction whose monotone curve the random-state result deviates from.","marker":"[4]"},{"why":"Defines the multi-entropy measure that the paper uses as its central entanglement diagnostic.","marker":"[14]"},{"why":"Provides the random tensor network replica method and large-bond-dimension approximation used for all computations.","marker":"[36]"},{"why":"Previous computation of Rényi multi-entropy in random tensor networks, which the paper compares with and extends to evaporating black hole curves.","marker":"[42]"},{"why":"Supplies the permutation-group and replica partition function techniques, including the 1/d corrections, used throughout.","marker":"[31]"}],"fun_headline_variants":["Multi-entropy curve peaks after Page time","Black hole multi-entropy leaves secret entanglement","Hawking radiation yields nonzero multi-entropy at end","Multipartite Page curve generalizes with later peak"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on treating the evaporating black hole plus all Hawking radiation as a single Haar-random pure state; a real evaporating black hole evolves unitarily, conserves energy, and produces radiation through semiclassical pair creation, so if the random-state approximation fails, the predicted peak time and nonzero endpoint need not match actual black hole radiation.","fun_headline_variants_meta":{"raw":{"variants":["Multi-entropy curve peaks after Page time","Black hole multi-entropy leaves secret entanglement","Hawking radiation yields nonzero multi-entropy at end","Multipartite Page curve generalizes with later peak"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000367,"raw_usage":{"total_tokens":1959,"prompt_tokens":923,"completion_tokens":1036,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":539,"completion_tokens_details":{"reasoning_tokens":986}},"tokens_in":539,"tokens_out":1036,"duration_ms":10192,"temperature":1.0,"reasoning_tokens":986,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T18:43:44.074547+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the q=3, n=2 multi-entropy in a concrete unitary evaporation model, such as a random unitary circuit or a spin-chain Hamiltonian that couples a black-hole register to radiation registers with fixed total dimension, and check whether $S_2^{{(3)}}$ peaks at dR=(dTotal)^{1/3} and remains nonzero at dBH=1; a peak at the Page time or a vanishing endpoint would disprove the paper's central claim.","supporting_citations":[],"review_version":1}