{"id":"899c6f13-8a36-423f-a12b-c19ea67e766b","arxiv_id":"2506.10064","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The entropy of a bulk region in holographic states equals the generalized entropy of the smallest wedge containing it, derived from a replica path integral via a hollow-graphic construction.","lead":"This paper derives the Bousso-Penington proposal for the entanglement wedge of a gravitating bulk region from the gravitational path integral, using random tensor networks and fixed-geometry states. The result gives a path-integral formula for the entropy of a bulk region and shows the final answer is independent of how the region is defined.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bulk-region diagonal approximation (Eq. 4.4) is not inherited from the boundary case unless the hollow-graphic map Va is an isometry; the paper only proves norm preservation and never proves off-diagonal holographic states are suppressed, so the n→1 limit may not uniquely yield BP.","rationale":"The paper aims to derive the Bousso-Penington proposal for the entropy of gravitating bulk regions from the gravitational path integral in time-reflection symmetric settings. The argument has two main legs: an RTN proof of the hollow-graphic construction, and a gravitational derivation using fixed-geometry states plus a diagonal approximation. The RTN leg is solid and gives a clear Hilbert-space interpretation, with the caveat that it is a toy model. The gravitational leg, however, rests on Eq. (4.4), which drops all h≠h' contributions to Tr(ρ_a^n). For boundary subregions, this diagonal approximation was established in Ref. [33]; for bulk subregions the paper does not prove the analogous statement. The reader's weakest-assumption analysis correctly identifies this as the key gap. The concern is load-bearing because if off-diagonal terms contribute, the n→1 limit need not select the BP surface: other saddles or superpositions of saddles could dominate, and the claimed universality of Eq. (4.9) would fail even though the individual fixed-geometry calculations are correct. The paper also explicitly acknowledges that the fixed-geometry basis is approximate due to wormhole corrections; this makes the transfer from boundary to bulk results non-automatic. The proposed concrete test directly checks whether the hollow-graphic map preserves inner products and whether off-diagonal contributions vanish in the semiclassical limit. The worked JT and minisuperspace examples are useful consistency checks, but they only evaluate the diagonal saddle and so cannot resolve this concern. The verdict should remain CONDITIONAL: the derivation is plausible and internally consistent under the stated assumption, but the missing justification of Eq. (4.4) for bulk regions should be supplied before the derivation is treated as fully general.","tokens_in":23596,"tokens_out":9980,"duration_ms":122756,"concrete_test":"Compute the off-diagonal overlap in JT gravity: for two fixed-dilaton states Φ1≠Φ2, evaluate Z_off = Tr_{∂Σ\\∂a}(|Φ1,a⟩⟨Φ2,a|) using the gravitational path integral with the slit (a±) boundary conditions, and compare with the diagonal norms. If Z_off is not exponentially suppressed relative to √(⟨Φ1,a|Φ1,a⟩⟨Φ2,a|Φ2,a⟩), or equivalently if ⟨Φ1,a|Φ2,a⟩≠⟨Φ1|Φ2⟩, then Va is not an isometry and Eq. (4.4) is unjustified. A stronger version: take a two-state superposition |ψ⟩=c1|Φ1⟩+c2|Φ2⟩ and compute Tr(ρ_a^n) including all off-diagonal terms; check whether the n→1 limit differs from Eq. (4.9). This can be done with the known two-boundary wormhole amplitude in JT gravity and the kink geometries of Sec. 5.1.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central derivation reduces to Eq. (4.4): after the linear extension (4.2), off-diagonal terms with h≠h' in Tr(ρ_a^n) are dropped 'motivated by replica symmetry'. For boundary subregions this was justified in Ref. [33], but for bulk regions the justification has to pass through the hollow-graphic map Va. The paper establishes only that the slit construction preserves the norm of each fixed-geometry state (⟨h_a|h_a⟩=⟨h|h⟩); it does not establish that Va is an isometry, i.e. that ⟨h_a|h'_a⟩=⟨h|h'⟩ for h≠h'. If Va is not isometric, the suppression of off-diagonal terms in the boundary calculation does not automatically transfer: the new boundaries ∂a are not traced over, so the mechanism that makes boundary off-diagonal overlaps small may be weaker. Moreover, the fixed-geometry basis is itself only approximate because of wormhole corrections, as the paper acknowledges in Sec. 3 (footnote 11) and Sec. 4. In JT gravity, fixed-dilaton states have non-zero wormhole overlaps; the hollowed overlaps could be comparable and contribute to Tr(ρ_a^n) at order e^{-c/G}. If such off-diagonal contributions survive in the semiclassical limit, Eq. (4.9) is not the unique n→1 limit and the claimed derivation of the BP proposal is incomplete. The examples in Sec. 5 only evaluate the diagonal saddle and therefore cannot detect this failure.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes a derivation of the Bousso-Penington (BP) proposal for generalized entanglement wedges of gravitating bulk regions using the gravitational path integral, in time-reflection symmetric settings. The authors first formulate the proposal in random tensor networks by 'hollowing out' a bulk region a, i.e., removing the tensors and internal edges in a and computing the entropy of the resulting open legs; this yields S(a)=A(γ_a)/(4G)+S(ρ_{E(a)}). They then use the connection between RTNs and fixed-geometry states in gravity, decompose a general holographic state as a superposition over fixed-geometry states, and invoke a diagonal approximation to compute Rényi entropies. They show that the saddles computing the Rényi entropies depend on how the bulk region is gauge-invariantly specified, but claim that the dependence drops out in the n→1 limit and that the BP proposal is universally recovered. Explicit computations are given in JT gravity for four different gauge-invariant specifications of an interval, and in Einstein gravity in a minisuperspace approximation for an annulus.","tokens_in":23941,"tokens_out":6245,"duration_ms":74729,"significance":"If the derivation is correct, this is an important result: it gives a gravitational path-integral rationale for the BP proposal and connects it to tensor-network constructions. The paper has several concrete strengths: a clear RTN derivation of the BP formula, a new notion of restricted entanglement wedges with a discussion of entropy inequalities, explicit worked examples in JT gravity and in a minisuperspace Einstein-gravity model, and a demonstration that the n>1 Rényi saddles depend on the gauge-invariant specification while the n→1 result does not. The main unresolved point is the bulk diagonal approximation, which is assumed rather than proven; this is a load-bearing step for the claimed derivation.","major_comments":[{"comment":"The diagonal approximation is the load-bearing step of the gravitational derivation, but it is not justified for bulk regions. For boundary subregions, the suppression of off-diagonal terms follows from the analysis of the fixed-geometry basis in Ref. [33]. Here, the hollowing map V_a is shown only to preserve the norm of each fixed-geometry state, ⟨h_a|h_a⟩=⟨h|h⟩, not to preserve the inner products ⟨h_a|h'_a⟩=⟨h|h'⟩ for h≠h'. Therefore the boundary-case argument does not automatically transfer: the new boundaries ∂a are not traced over in the same way as the boundary complement, and one must show that off-diagonal hollowed overlaps are suppressed at order e^{-c/G}. The examples in Sec. 5 evaluate only the diagonal saddle and cannot detect such contributions. Please provide an estimate or proof of off-diagonal suppression in the hollowed Hilbert space, or state this as an explicit assumption with its consequences.","section":"Sec. 4, Eq. (4.4)"},{"comment":"The claimed universality of the n→1 limit depends on the assumption that the gauge-invariant prescription used to define a on off-shell geometries does not affect the limit. The paper argues that all prescriptions agree on the leading geometry h_ψ, so h_1=h_ψ; however, the off-diagonal terms in Eq. (4.3), if not negligible, can depend on the extension of a to geometries h≠h', and the n→1 limit of the full expression need not be independent of the prescription. Please show that off-diagonal contributions are subleading before taking n→1, or give a separate argument that the n→1 limit commutes with the diagonal approximation.","section":"Sec. 4, after Eq. (4.9)"},{"comment":"The paper acknowledges, in Sec. 3 (footnote 11) and Sec. 4 after Eq. (4.2), that the fixed-geometry states form only an approximate basis because of wormhole corrections. This is particularly important for the hollowing construction, since wormhole overlaps in the hollowed Hilbert space could be of the same order as the diagonal contributions to Tr(ρ_a^n). In JT gravity, fixed-dilaton states have non-zero wormhole overlaps; the examples in Sec. 5 do not compute these. Please quantify the size of wormhole corrections to the hollowed inner products in the regimes considered, or explain why they are subleading in the n→1 limit.","section":"Sec. 3 and Sec. 4"}],"minor_comments":[{"comment":"The factor (m−1) in Eq. (B.6) should presumably be (n−1); the symbol m is not defined and the surrounding equations use n.","section":"Appendix B, Eq. (B.6)"},{"comment":"The displayed limit contains the stray notation 'ℓ − − − →'; this should be a single arrow or phrased as a limit.","section":"Sec. 5.1, Eq. (5.9)"},{"comment":"Reference [36] appears to duplicate Reference [32]; the two entries for the modified cosmic brane proposal should be consolidated.","section":"References"},{"comment":"The caption contains the phrase 'null like', which should be 'null'.","section":"Sec. 6.2, Fig. 12 caption"},{"comment":"The density matrix ρ_a is not manifestly normalized; please state the normalization convention used in the Rényi trace, as is done for the boundary case in Eq. (3.5).","section":"Sec. 4, after Eq. (4.3)"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of hep-th and is likely to be of interest to the journal's readership. The main risk is the unproven bulk diagonal approximation, which is central to the claimed derivation of the BP proposal. The RTN construction and the explicit examples provide strong supporting evidence, so I do not see grounds for rejection; however, the authors should either prove or carefully quantify the off-diagonal and wormhole corrections in the hollowed setting before the derivation can be considered complete."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is not another restatement of BP. It gives a concrete random-tensor-network construction—hollowing out a bulk region and computing the entropy of the opened legs—and then maps that construction onto a gravitational path integral using fixed-geometry states. The restricted entropy inequalities and the observation that the Rényi saddles depend on the gauge-invariant specification of the region are new and useful, and the JT and minisuperspace examples are worked out in enough detail to be checked.\n\nWhat is solid: the RTN part is rigorous in the large-bond-dimension limit, and it cleanly reproduces the BP entanglement wedge for bulk regions. The fixed-geometry technology is used carefully, and the paper is honest about the approximate nature of the fixed-geometry basis and about the difference between its construction and earlier related work in Refs. [12], [16], [38]. The examples showing that the n→1 limit is independent of how the region is gauge-invariantly specified are concrete and plausible.\n\nThe load-bearing step is Eq. (4.4), the diagonal approximation for bulk regions. For boundary regions this was justified in [33]. For bulk regions, the justification would have to pass through the hollow-graphic map Va. The paper proves that Va preserves norms, but it does not prove that Va is an isometry, and nothing in Secs. 4 or 5 shows that off-diagonal fixed-geometry overlaps are suppressed after hollowing. The examples only evaluate the diagonal saddle, so they cannot detect a failure of this assumption. This makes the derivation conditional rather than wrong. The paper explicitly says the diagonal approximation is motivated by replica symmetry, and for sharply peaked semiclassical states one might expect such suppression, but it remains an assumption—and it is the same assumption that does the work in reaching Eq. (4.9). The stress-test note is therefore on target: unless Va is isometric, the boundary-region suppression mechanism does not automatically transfer.\n\nA second soft spot is that gamma_a in Eq. (4.6) is already defined as the surface homologous to partial a that minimizes the generalized entropy of E(a). So Eq. (4.9) has some circularity built in; the independent content comes from the RTN calculation and the explicit saddle-point examples. I would not call this fatal, but the paper could be more upfront that the minimization is encoded in the definition.\n\nNet: this is a serious, worthwhile paper. The derivation is important enough that the bulk-region diagonal approximation should be either proven or clearly marked as a conjecture. I would send it to a serious referee, and I would cite it in my own work with a qualifier.","headline":"A genuinely new path-integral derivation of the Bousso-Penington proposal for bulk regions, worth serious referee time, but the central diagonal approximation is assumed rather than proven.","tokens_in":24455,"tokens_out":2054,"would_cite":true,"duration_ms":26482,"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":"Path integral derivation shows a bulk region's entropy is the generalized entropy of its smallest enclosing entanglement wedge.","keywords":["holographic entanglement entropy","entanglement wedge","gravitational path integral","random tensor networks","generalized entropy","replica trick","Renyi entropy","Bousso-Penington proposal"],"falsifier":"Find a static holographic state and a gauge-invariant bulk region where replica symmetry breaking or off-diagonal fixed-geometry contributions dominate the Rényi entropy, and show the n→1 limit deviates from A(γ_a)/(4G) + S(ρ_{E(a)}). A concrete example would be a state with competing extremal surfaces of nearly equal area, where the diagonal approximation is known to fail for boundary regions, used as input for a bulk region.","tokens_in":23377,"feed_emoji":"","tokens_out":2686,"duration_ms":22565,"temperature":0.7,"pith_summary":"This paper claims that the Bousso-Penington (BP) proposal for the entanglement wedge of a gravitating bulk region follows from the gravitational path integral in time-reflection symmetric settings. The authors define the entropy of a bulk region by a hollow-graphic construction—removing the region to create new boundaries—and compute its Rényi entropy through fixed-geometry states. They show that in the n→1 limit the result equals the generalized entropy A(γ_a, h_ψ)/(4G) + S(ρ_{E(a)}), independent of how the bulk region is gauge-invariantly specified. This matters because it elevates the BP proposal from a geometric conjecture to a consequence of the gravitational path integral.","feed_headline":"Path integral yields the BP rule for bulk entanglement wedges","feed_subtitle":"Removing a gravitating region and sewing up the cut reproduces its generalized entropy at n=1, in any gauge-invariant description.","key_machinery":"The hollow-graphic state |ψ_a⟩: the state obtained by removing (hollowing out) the bulk region a, generating new open boundaries a±, and computing entropies of the resulting open legs. Combined with fixed-geometry states |h⟩, it reduces the bulk-region entropy computation to a random-tensor-network-style replica calculation where the only nontrivial saddles are two domains separated by a domain wall, whose area gives the generalized entropy.","core_discovery":"The central claim is that for a gravitating bulk region a in a time-reflection symmetric holographic state, the von Neumann entropy S(a) computed from the hollow-graphic replica path integral equals the generalized entropy of the smallest region E(a) that contains a and shares its conformal boundary: S(a) = A(γ_a, h_ψ)/(4G) + S(ρ_{E(a)}). The paper derives this by expanding the state in fixed-geometry states, applying the hollow-graphic map, computing Rényi entropies via replica saddles with a diagonal approximation, and taking n→1. The n-dependence of the saddles depends on the gauge-invariant specification of a, but the dependence drops out in the n→1 limit, so the BP proposal is recovered universally.","pith_inferences":["The proof's reliance on the diagonal approximation suggests that bulk-region entropies in states with strong replica symmetry breaking may deviate from the BP formula; testing such states in JT gravity is a concrete next step.","The gauge-invariance of the n→1 limit hints that the BP formula is a robust semiclassical statement, but the n-dependent saddles imply that Rényi entropies of bulk regions are not universal—they encode how the region is defined.","The restricted-wedge formulation points toward a fully holographic proof of the entropy cone for gravitating regions that does not require the stronger independence conditions used in earlier work.","A Lorentzian derivation would be needed to extend the hollow-graphic construction to time-dependent settings; the paper's static Euclidean method cannot capture the time-dependent BP proposal."],"forward_implications":["If the derivation holds, the BP proposal is not an additional postulate but a consequence of the gravitational path integral in static settings.","The generalized entanglement wedge of a bulk region is independent of how the region is gauge-invariantly prescribed, resolving potential ambiguities in defining bulk subregions.","Restricted entanglement wedges (computed in a spacetime with other regions treated as boundaries) provide the natural objects for entropy inequalities, unifying SSA and the holographic entropy cone.","The same hollow-graphic method extends to compute min and max entanglement wedges via Petz-map reconstruction for incompressible states where replica symmetry breaks."],"supporting_citations":[{"why":"Bousso-Penington proposal that the paper aims to derive.","marker":"[10]"},{"why":"Time-dependent extension of the BP proposal and the independence condition used for SSA.","marker":"[11]"},{"why":"Random tensor network formalism providing the Ising-like replica computation.","marker":"[20]"},{"why":"Diagonal approximation for holographic Rényi entropies, the key step carried over to bulk regions.","marker":"[33]"},{"why":"Lewkowycz-Maldacena derivation that the paper generalizes for bulk regions.","marker":"[34]"},{"why":"Fixed-area states and replica symmetry breaking, used for min/max wedges and clarifying when the diagonal approximation fails.","marker":"[31]"},{"why":"Replica wormhole computations and Petz map reconstruction used in deriving RTN/gauge-gravity connection and min/max wedges.","marker":"[6]"},{"why":"Modified cosmic brane proposal showing the diagonal approximation gives correct n<1 Rényi entropy.","marker":"[36]"}],"fun_headline_variants":["Path integral derivation of BP entanglement wedges","Hollow-gravity proof of generalized entanglement wedge entropy","Hollow-grams: path integral yields BP wedge rule","Gravitational path integral recovers BP entanglement proposal","From hollowed tensors to BP generalized entropy"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The diagonal approximation: off-diagonal terms in the fixed-geometry expansion are dropped, motivated by replica symmetry, without a proof that this holds for bulk regions defined by the hollow-graphic map.","fun_headline_variants_meta":{"raw":{"variants":["Path integral derivation of BP entanglement wedges","Hollow-gravity proof of generalized entanglement wedge entropy","Hollow-grams: path integral yields BP wedge rule","Gravitational path integral recovers BP entanglement proposal","From hollowed tensors to BP generalized entropy"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000181,"raw_usage":{"total_tokens":1285,"prompt_tokens":898,"completion_tokens":387,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":514,"completion_tokens_details":{"reasoning_tokens":313}},"tokens_in":514,"tokens_out":387,"duration_ms":5066,"temperature":1.0,"reasoning_tokens":313,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T04:35:54.122618+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a static holographic state and a gauge-invariant bulk region where replica symmetry breaking or off-diagonal fixed-geometry contributions dominate the Rényi entropy, and show the n→1 limit deviates from A(γ_a)/(4G) + S(ρ_{E(a)}). A concrete example would be a state with competing extremal surfaces of nearly equal area, where the diagonal approximation is known to fail for boundary regions, used as input for a bulk region.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Modified cosmic brane proposal showing the diagonal approximation gives correct n<1 Rényi entropy."}],"review_version":1}