{"id":"9d5ec2df-cff0-4412-a115-434d7da9a774","arxiv_id":"1908.02297","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Finite differences between holographic entanglement entropies are independent of the radial cutoff, making the globally minimal surface prescription well-posed.","lead":"This paper proves that a standard way to tame infinite areas in holographic entropy calculations gives the same answer no matter which radial coordinate is used for the cutoff. It shows the 'choose the smallest surface' rule and finite quantities like mutual information are well-defined, and says exactly when vacuum subtraction works.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Smoothness of extremal surfaces near the AdS boundary is the key unproven premise; the order-by-order induction fails for non-smooth surfaces, leaving the theorem conditional.","rationale":"After reviewing the proof, the central argument is internally consistent under its stated assumptions: the order-by-order induction in Section 3.1, with Lemma 2 and eq. (3.16), does establish universality of the binormal to order z^{d-3} and of the surface position to order z^{d-2}, and Section 3.2 correctly uses this to prove cutoff-independence of finite area differences. The only unproved premise that could invalidate the central claim is the smoothness of the extremal surfaces near the boundary, exactly the reader's weakest assumption. Conjecture 3 affects only the secondary equivalence of direct versus transported cutoffs, not the two main results. The paper explicitly acknowledges the smoothness assumption and the conjecture, and no internal inconsistency or external counterexample was found. Therefore the reader's CONDITIONAL verdict is appropriate, and no verdict change is warranted.","tokens_in":33087,"tokens_out":49066,"duration_ms":533304,"concrete_test":"Take a boundary-anchored extremal surface that is only C^k (k < infinity) near the boundary, for example by imposing a C^2 but non-C^3 embedding in a smooth asymptotically AdS background, and numerically compute the area difference between the boundary-anchored and transported cutoff prescriptions for two different defining functions. If the finite area difference changes with the defining function, the smoothness assumption is load-bearing. Alternatively, prove or disprove elliptic boundary regularity for HRT surfaces in smooth, globally hyperbolic asymptotically AdS spacetimes: if such surfaces are always smooth up to the conformal boundary for smooth entangling surfaces, the smoothness concern is neutralized.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that homologous extremal surfaces have identical area divergences and cutoff-independent finite differences rests on the Section 3.1 induction, which expands the surface position f and the smooth binormal ~N in Taylor series in the defining function z (eqs. 3.13-3.16 and Lemma 2). This requires the surface to be smooth in a neighborhood of the boundary, as explicitly assumed in Section 1. If an extremal surface develops a caustic, cusp, or other non-smooth feature near the boundary, the expansions and the inductive solution of the binormal divergence equation break down. Since the paper does not prove that HRT surfaces always possess this regularity, the theorem is conditional on an analytic/regularity premise that is not established. The local integrability of the normal bundle is a secondary assumption, but Appendix B argues it can be relaxed, so it is less load-bearing. The same smoothness assumption also underlies the computation of the area integrals in Section 3.2, so the cutoff-independence proof inherits the limitation. Thus the reader's weakest-assumption identification is accurate and is the main barrier to an unconditional central claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper gives a careful treatment of radial cutoffs for holographic entanglement entropy in asymptotically AdS spacetimes. The author introduces a notion of cutoff based on arbitrary defining functions and proves, under stated smoothness and normal-bundle integrability assumptions, that any two extremal surfaces anchored to the same boundary region have identical area divergences and that the finite difference of their regulated areas is independent of the choice of defining function. The main consequences are that the \"globally minimal surface\" prescription for the HRT surface is regulator-independent and that divergence-cancelling combinations such as mutual information are well defined. The paper also analyzes vacuum subtraction under a relaxed Fefferman-Graham falloff and, subject to Conjecture 3, argues that direct and transported cutoff prescriptions give the same finite area differences. Appendices provide a generalized Fefferman-Graham expansion and a covariant toolkit for extremal surfaces based on the divergence of the unit binormal.","tokens_in":33312,"tokens_out":13380,"duration_ms":143561,"significance":"If the central result holds, it fills a real gap in the holographic literature: the cutoff-independence of HRT areas is frequently assumed but, to my knowledge, has not been proved at this level of detail. The paper's main technical contribution, the order-by-order solution of the binormal divergence equation in Section 3.1, is original and carefully executed, and the covariant formalism in Appendix B is clean and likely to be useful beyond this application. The paper is also transparent about its scope: the smoothness assumption, the local integrability of the normal bundle, and Conjecture 3 are all stated explicitly. The weakness is that the central theorem is conditional on a regularity premise that is not proved or cited, while the title and abstract advertise the unconditional statement. This is a scope issue rather than a circularity or internal-inconsistency issue; the paper does not assume the result it proves.","major_comments":[{"comment":"The central theorem is conditional on the smoothness assumption stated in Section 1, but the title and abstract assert cutoff covariance without that qualifier. The induction proving universality of the binormal up to order z^{d-3} requires f1, f2, and ~N^{\\mu\\nu} to admit Taylor expansions in the defining function z; Lemma 2 uses this expansion at every order, and the same regularity is needed for the area integrals in Section 3.2. The paper neither proves nor cites a regularity theorem ensuring that extremal surfaces anchored on a smooth boundary region are smooth in a neighborhood of the AdS boundary. Please either prove or cite such a regularity result, or reformulate the main theorem and the abstract as applying to smooth extremal surfaces and explicitly state that non-smooth surfaces are outside the scope of the proof. This is the main obstacle to an unconditional version of the claimed result.","section":"Section 1 and Section 3.1 (Lemma 2, Eqs. (3.13)–(3.16))"}],"minor_comments":[{"comment":"There is a typo: the coordinate range should be x3, ..., x_{d-1}, not x_{d1}.","section":"Section 3.2, text after Eq. (3.38)"},{"comment":"The notation N_{\\mu\\nu} = z^2 N_{\\mu\\nu} reuses the symbol N for the smooth down-index binormal; this is confusing because N already denotes the physical binormal. Please use a different symbol, e.g. \\tilde N_{\\mu\\nu}, for the rescaled down-index object.","section":"Section 3.1, Eq. (3.28) and surrounding text"},{"comment":"The equivalence of the direct and transported cutoff prescriptions depends on unproved Conjecture 3 and is not a theorem. The sentence \"we conclude that these two prescriptions give equivalent answers\" should be preceded by an explicit conditional clause, e.g. \"Assuming Conjecture 3, we conclude...\", so that the conjecture-dependence of this corollary is not lost on the reader.","section":"Section 3.2, Conjecture 3 and Eqs. (3.40)–(3.42)"},{"comment":"The argument that the change in the finite piece under a change of defining function is determined by divergent terms is correct, but it would be clearer to point out explicitly that the coefficient of the z^{-1} term in the integrand of Eq. (3.38) is universal by the results of Section 3.1, and that this coefficient is the only source of a finite change in the finite piece. Adding one sentence would make the logic of the cutoff-independence proof easier to follow.","section":"Section 3.2, discussion of the cutoff change"},{"comment":"The proof is formulated locally in coordinates where x1 and x2 are non-tangent to the extremal surface, and the paper states that the argument can be patched over an open cover of the entangling surface. Since this patching is invoked but suppressed, a brief remark in the main text describing how the local results combine would strengthen the rigor of the presentation.","section":"Section 1, footnote 12"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid proof of a folklore result under explicit regularity assumptions; I see no circularity, no fitted parameters, and no reason to doubt the internal consistency of the derivation. The main issue is that the title, abstract, and introduction claim more than the theorem actually establishes unless the smoothness assumption is either proved or explicitly incorporated into every statement of the results. This is fixable by revision: either prove or cite a boundary regularity theorem for extremal surfaces, or systematically qualify the claims. I would not reject the paper on these grounds."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nYou should know about this paper because it settles a question most of us had stopped asking: whether the radial cutoff prescription for HRT surface areas is actually coordinate-independent. Sorce proves that two extremal surfaces homologous to the same boundary region have identical area divergences in any defining-function cutoff, and the finite difference between them is cutoff-invariant. That means the 'globally minimal surface' rule is well-posed, and divergence-cancelling quantities like mutual information are well-defined. The literature routinely assumed these facts; the proofs were not there.\n\nWhat is new is the machinery. The covariant analysis of extremal surfaces in Appendix B, including the k-normal-form divergence condition and four equivalent characterizations of extremality, is a genuinely useful toolkit. Two of those characterizations are presented as original. The generalized Fefferman-Graham theorem in Appendix A, allowing arbitrary matter falloff, is also a clean extension. The paper is honest about what it assumes: smoothness of the extremal surfaces near the boundary, local integrability of the normal bundle, and Conjecture 3 for the equivalence of two cutoff recipes. The central claims do not rely on Conjecture 3.\n\nThe stress-test concern lands, but only partly. The induction in Section 3.1 requires the surface to admit a Taylor expansion in the defining function all the way to the boundary. If an extremal surface develops a caustic or cusp near the boundary, the proof does not apply. The paper states this assumption explicitly in Section 1 but does not prove that HRT surfaces always satisfy it. So the main theorem is conditional on a regularity premise. That is a genuine scope limitation, but it is explicit and it matches assumptions used broadly in the literature. The reader's CONDITIONAL verdict is right; this is not a hidden flaw, just a clearly marked boundary of the result.\n\nThe secondary equivalence between 'direct' and 'transported' cutoffs depends on Conjecture 3, which is labeled as a conjecture and used only for that corollary. If you care about the main result, you can ignore it.\n\nOverall: this is a solid, careful paper that deserves a serious referee. The physics was assumed, but the rigorous cleanup is exactly what the holographic entanglement literature needs. I would cite it in my own work on cutoff dependence, and I would bring it to a reading group that cares about the foundations of HRT. Send it to peer review; the referee should check the induction and the Appendix B identities, and should probe the smoothness assumption. Not a desk reject.","headline":"A rigorous proof that the HRT area-difference cutoff is independent, grounded in an explicit smoothness assumption; the physics was assumed, the proofs are new and worth having.","tokens_in":33801,"tokens_out":2949,"would_cite":true,"duration_ms":31642,"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":"This paper proves that any two extremal surfaces in an asymptotically AdS spacetime anchored to the same boundary region have identical area divergences under every smooth radial cutoff, and that their finite area difference is…","keywords":["holographic entanglement entropy","extremal surfaces","AdS/CFT correspondence","cutoff independence","conformal infinity","Fefferman-Graham expansion","vacuum subtraction","mutual information"],"falsifier":"Find an asymptotically AdS spacetime with two candidate extremal surfaces for the same boundary region, regulate their areas with two different defining functions $z$ and $e^{\\omega}z$, and check whether the $z_c^0$ term of their area difference is identical; a single case where it changes would falsify the paper's central claim.","tokens_in":32864,"feed_emoji":"🌌","tokens_out":9707,"duration_ms":99806,"temperature":0.7,"pith_summary":"The paper establishes that radial-cutoff regulation of holographic entanglement entropy is covariant: in an asymptotically AdS spacetime, any two extremal surfaces that approach the same boundary region have identical area divergences for every choice of defining function $z$, and the finite difference between their areas is independent of which defining function is chosen. This makes the \"globally minimal surface\" rule pick out the same Hubeny-Rangamani-Takayanagi surface in every radial coordinate, so the entanglement wedge of a boundary region is determined by the region itself and not by a choice of regulator. It also guarantees that divergence-cancelling CFT quantities such as mutual information are well-defined under the holographic dictionary. The proof solves the covariant binormal divergence equation order by order in $z$, showing the surface's binormal is fixed by boundary data up to order $z^{d-3}$ and its coordinate position up to order $z^{d-2}$. The paper additionally identifies when vacuum subtraction between different spacetimes is valid: spacetimes obeying the relaxed Fefferman-Graham falloff $R_{ab}=-(d-1)g_{ab}+o(z^{d-5})$ have universal, state-independent divergences.","feed_headline":"Radial cutoff choice cannot change the holographic entropy difference","feed_subtitle":"Holographic entanglement entropy's finite parts and minimal-surface selection survive any smooth radial cutoff.","key_machinery":"The machinery is the covariant characterization of extremality by the vanishing divergence of the unit binormal. A codimension-2 surface $\\Sigma$ has a unique unit binormal $N_{ab}$ normalized by $N_{ab}N^{ab}=-2$; $\\Sigma$ is extremal iff $\\nabla_a N^{ab}|_\\Sigma=0$, independent of how $N_{ab}$ is extended off the surface. Writing this equation in coordinates adapted to a defining function $z$, with rescaled binormal $\\tilde N^{\\mu\\nu}=N^{\\mu\\nu}/z^2$, produces component equations (3.9)--(3.10) that are solved inductively near $z=0$. Lemma 2 supplies the inductive step: the binormal through order $z^n$ fixes the surface's coordinate position through $z^{n+1}$, and that position fixes the tangential binormal at $z^{n+1}$; the induction runs until order $z^{d-3}$, where the coefficient $(d-2-n)$ vanishes. This yields universal divergent structure, and the cutoff-independence of finite differences follows by comparing the area between two cutoff surfaces, where only the divergent terms matter. The equivalence between direct and transported cutoff prescriptions uses the same divergence control plus Stokes' theorem on an extremal foliation, assuming Conjecture 3.","core_discovery":"The central claim is that the finite difference in area between any two extremal surfaces homologous to the same boundary region is cutoff-independent. More precisely, in any asymptotically (locally) AdS spacetime and for any defining function $z$, two extremal surfaces sharing the same boundary anchor have the same divergent area terms under the cutoff $z=z_c$, and the $z_c^0$ term in their area difference does not change when $z$ is replaced by another defining function $e^{\\omega}z$. Consequently the \"globally minimal surface\" among several extremal candidates is well-defined: unless two candidates are exactly degenerate, exactly one candidate has a negative finite area difference with every other candidate, and that ordering is independent of the regulator. The same argument shows that cutting off the boundary-anchored surface directly and instead anchoring an extremal surface to the transported cutoff region give the same finite area differences, subject to Conjecture 3 on the existence of an extremal foliation. Under the relaxed Fefferman-Graham falloff, area divergences are universal across spacetimes, so vacuum-subtracted entanglement entropy is well-defined; spacetimes with slower matter falloff evade this and can have state-dependent divergences.","pith_inferences":["Editorial extension: the same order-by-order binormal expansion predicts a practical numerical test: compute HRT surface area differences for a nontrivial boundary region using two independent defining functions; the $z_c^0$ term should agree to all orders, and any disagreement would localize a broken assumption.","Editorial extension: the local, covariant proof should apply to other boundary-anchored extremal objects, such as subleading extremal saddles or entwinement probes, because the divergence control does not require global minimality.","Editorial extension: Conjecture 3, not the main theorem, is the pressure point for the direct-versus-transported equivalence; in dimensions $d\\ge 5$ the area difference between the two prescriptions can diverge for a single surface, so a counterexample to the conjecture would make the corollary fail even though the main theorem stands."],"forward_implications":["The choice of radial coordinate cannot change which extremal surface is globally minimal, so a boundary region's HRT surface and entanglement wedge are determined by the region alone.","Mutual information and conditional mutual information, computed as divergence-cancelling sums of surface areas, are finite and cutoff-independent in every asymptotically AdS spacetime.","Vacuum subtraction is legitimate exactly for spacetimes satisfying the relaxed Fefferman-Graham falloff; in slower-falloff spacetimes the divergences are state-dependent and the vacuum-subtracted entropy remains infinite.","Directly cutting off a boundary-anchored surface and finding a new extremal surface anchored to the transported cutoff region give the same finite area differences, provided Conjecture 3 holds."],"supporting_citations":[{"why":"Introduces the Ryu-Takayanagi holographic entanglement entropy proposal whose radial regulation is the subject of the paper.","marker":"[1]"},{"why":"Proposes the covariant HRT prescription and the \"globally minimal surface\" rule that the paper proves is cutoff-independent.","marker":"[2]"},{"why":"Defines asymptotically locally AdS spacetimes, the class to which the paper extends its main results.","marker":"[7]"},{"why":"Identifies state-dependent divergences in holographic entanglement entropy; the paper's vacuum-subtraction analysis targets this phenomenon.","marker":"[8]"},{"why":"Supplies the conformal-infinity and Fefferman-Graham asymptotic-expansion framework underlying defining functions and the cutoff construction.","marker":"[15]"},{"why":"Proves existence and uniqueness of special defining functions, used when comparing regulators between different spacetimes.","marker":"[16]"}],"fun_headline_variants":["Holographic entropy finite parts are cutoff-covariant","Radial cutoff doesn't shift minimal surface selection","Cutoff-independent entanglement entropy in AdS/CFT","Finite entropy differences survive any radial regulator","Minimal surface and finite entropy: cutoff-independent"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes every boundary-anchored extremal surface is smooth in a neighborhood of the AdS boundary, so that its binormal and coordinate position admit Taylor expansions in the defining function all the way to $z=0$; a caustic, cusp, or other nonsmooth feature would put the universal-divergence and cutoff-independence arguments outside their stated domain.","fun_headline_variants_meta":{"raw":{"variants":["Holographic entropy finite parts are cutoff-covariant","Radial cutoff doesn't shift minimal surface selection","Cutoff-independent entanglement entropy in AdS/CFT","Finite entropy differences survive any radial regulator","Minimal surface and finite entropy: cutoff-independent"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000485,"raw_usage":{"total_tokens":2476,"prompt_tokens":1108,"completion_tokens":1368,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":724,"completion_tokens_details":{"reasoning_tokens":1294}},"tokens_in":724,"tokens_out":1368,"duration_ms":14314,"temperature":1.0,"reasoning_tokens":1294,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:48:36.471988+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find an asymptotically AdS spacetime with two candidate extremal surfaces for the same boundary region, regulate their areas with two different defining functions $z$ and $e^{\\omega}z$, and check whether the $z_c^0$ term of their area difference is identical; a single case where it changes would falsify the paper's central claim.","supporting_citations":[{"cited_title":"Fefferman and C","cited_arxiv_id":null,"evidence_quote":"Supplies the conformal-infinity and Fefferman-Graham asymptotic-expansion framework underlying defining functions and the cutoff construction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proves existence and uniqueness of special defining functions, used when comparing regulators between different spacetimes."}],"review_version":1}