{"id":"0d7a82cf-6d61-4cf0-ac8d-939ce4555601","arxiv_id":"1312.5510","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Holographic Rényi entropy for two intervals in a CFT with W symmetry matches the gravity computation at classical and 1-loop level to order x^8.","lead":"Physicists computed the same entanglement quantity two ways, once in a 3D gravity theory with higher spin fields and once in the dual quantum field theory. The two answers agree exactly through high order in a small parameter, a check of the holographic dictionary for higher spin gravity.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"No significant objection identified: the 1-loop gravity/CFT match is an independent, nontrivial check, and the acknowledged limitations (classical saddle assumption, no gravity comparison at 2-loop and 3-loop, overlap with [35]) do not undermine the central claim.","rationale":"The reader's weakest assumption (the classical saddle is unchanged by higher spin fields) is correctly identified as an assumption, but it is not the load-bearing part of the argument: in the ground state the W charges vanish, so the W fields contribute first at O(c^0), and the classical O(c) part is trivially the ordinary CFT result. The genuinely load-bearing check is the 1-loop match, which the paper performs through x^8. I verified one coefficient explicitly (n=2, x^6) and it matches exactly. The gravity 1-loop formula (2.7) is cited from the published [39], and the CFT operator content is enumerated level by level in Appendix A with d_K computed in Appendix B, so the match is not circular. The acknowledged limitations (footnote 1, Section 4's remarks on 2-loop and 3-loop, the overlap with [35]) are stated honestly and do not undermine the central claim. The abstract's 'confirms' is a mild overclaim for a single-observable check, but this is a presentational issue, not a correctness risk. The paper substantively advances the higher spin holography subfield, and ACCEPT with moderate confidence is appropriate.","tokens_in":29459,"tokens_out":21999,"duration_ms":407785,"concrete_test":"Independently recompute the x^8 coefficient of I^{(2,3,4),1-loop}_n from the OPE sum in Section 3.5 (using the d_K in Appendix B and the summation formulas in Appendix E) and compare it to the sum of (3.16), (2.8), and (2.9). If the three agree term by term in n, the central claim is confirmed at the highest order computed; if a discrepancy appears, the completeness of the quasiprimary list or the gravity 1-loop formula (2.7) for higher genus would be called into question.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that the 1-loop Rényi mutual information computed in higher spin AdS3 gravity matches the CFT computation from the twist-operator OPE including W3 and W4 quasiprimaries, through order x^8. For this to hold, two things must be true: (i) the gravity 1-loop partition function (2.7) must be the correct handlebody result for higher genus, and (ii) the CFT OPE sum must be complete through level 8. Both are checked by the match itself: the gravity formula is cited from [39] (a published source) and the CFT operator list is enumerated level by level in Appendix A with d_K computed in Appendix B. I verified the n=2, x^6 coefficient explicitly: the CFT difference I^{(2,3),1-loop}_2 - I^{(2),1-loop}_2 from (3.20) and (3.16) equals 1.192e-7, and the gravity spin-3 result (2.8) gives 2/(4096·4096) = 1.192e-7, an exact match. The reader's weakest assumption (the classical handlebody saddle is unchanged by higher spin fields) is valid in the ground state because the W charges vanish, so the W fields contribute first at O(c^0) (level 6 for W3^2, level 8 for W4^2), consistent with the gravity side where they are massless 1-loop fluctuations. The classical part is therefore trivially the ordinary CFT result, and the nontrivial content is the 1-loop match, which is independently verified. The paper's acknowledged limitations (footnote 1 citing [35] for the classical saddle, Section 4 noting no gravity comparison at 2-loop and 3-loop, and the stated overlap with [35]) are honest and do not affect the central claim. The abstract's 'confirms' is a mild overclaim for a single-observable check, but not a correctness concern.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The paper computes the Rényi mutual information for two disjoint intervals with small cross ratio x in a 2D CFT with W symmetry, both holographically (higher spin AdS3 gravity) and directly (twist-operator OPE), through order x^8. The classical and 1-loop results match exactly, and the paper also reports 2-loop and 3-loop CFT results. The stated conclusion is that the higher spin gravity/CFT correspondence is confirmed at 1-loop for the ground state, and that the holographic computation of Rényi entropy, including quantum corrections, is supported in both pure and higher spin AdS3 gravity.","tokens_in":29862,"tokens_out":10416,"duration_ms":186494,"significance":"The paper provides a nontrivial check of the higher spin AdS3/CFT2 correspondence beyond the genus-1 partition function. The explicit order-by-order computation to x^8 is a real check: the CFT side enumerates quasiprimary operators level by level in Appendix A and computes d_K coefficients in Appendix B, while the gravity side uses the 1-loop partition function (2.7) taken from [39]. The 2-loop and 3-loop CFT results (3.17), (3.18) are new and provide a target for future gravity comparisons. The paper also gives a useful byproduct in Appendix D on the one-interval cylinder case.","major_comments":[{"comment":"The paper assumes that the classical handlebody geometry is unchanged by the presence of higher spin fields, so that the classical Rényi entropy is still given by the Ryu-Takayanagi area formula. This assumption is acknowledged and cited to [35], but the paper's own justification is thin: it states that 'all the higher spin fields could be set to vanish without spoiling the underlying correspondence' without making explicit why the W charges vanish in the ground state and why this makes the classical part trivially the ordinary CFT result. Since the abstract claims a 'classical match' between gravity and CFT, the paper should state clearly that the classical match is trivial (both sides reduce to the ordinary CFT result) and that the nontrivial content is the 1-loop match. This is a framing issue rather than a correctness issue, but it is load-bearing for the abstract's claim.","section":"Section 2, footnote 1","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The sentence describing I^(2,3,4) is garbled: it mixes the 3-loop statement with the 1-loop statement, so the reader cannot tell which parts of I^(2,3,4) equal which parts of the earlier results. The sentence should be rewritten to state separately that I^(2,3,4)tree = I^(2)tree, I^(2,3,4)2-loop = I^(2,3)2-loop, I^(2,3,4)3-loop = I^(2)3-loop, and I^(2,3,4)1-loop = I^(2)1-loop + I^{spin-3,1-loop} + I^{spin-4,1-loop}.","section":"Section 3.5, Eq. (3.24)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"There is a typo '1-1oop' in the text introducing Eq. (3.16); it should read '1-loop'.","section":"Section 3.2, Eq. (3.16)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The error term is written 'O(x)^9'; it should be 'O(x^9)'. The same typo appears in Eq. (3.18).","section":"Section 2.2.2, Eq. (2.8)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The W4 normalization α_{W4} = α_{W4W4} = 1 is a choice. The paper should state explicitly that the product α_K d_K^2 is normalization-independent, so the match with the gravity result is unaffected by this choice.","section":"Appendix A, Eq. (A.22)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The claim that operators such as W3_{j1}W4_{j2} at level 7 have vanishing d_K is asserted without derivation. The paper should provide a check or a reference showing that these mixed operators do not contribute to the OPE of twist operators.","section":"Section 3.3, paragraph after Eq. (A.20)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phrase 'This confirms the higher spin gravity/CFT correspondence' is too strong for a 1-loop ground-state check. The paper should say 'provides a nontrivial check of' rather than 'confirms'.","section":"Abstract","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The statement that 'there is weak sign that the m-loop correction with m ≥ 3 is vanishing for S_n, n = 2, 3, ..., m' is speculative. The paper should be clearer that this is a conjecture based on the n=3, 3-loop observation, not a derived result.","section":"Section 4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The paper states that 'There is no further contributions from the W-generated operators to the tree-level result' and attributes this to the absence of higher spin contributions to the classical action. It would help to state explicitly that this is because the W_j at level 3 has vanishing d_K, so the W fields first contribute at O(c^0).","section":"Section 3.3, first bullet point","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"The paper discloses overlap with [35] (Perlmutter) in a 'Note added', which is good practice. The results are consistent with that work. The technical computation is detailed and explicit, and the central 1-loop match is verified by the reader's take at n=2, x^6. The remaining issues are local presentation and framing points, not correctness concerns."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"Worth your time if you work on higher spin holography or short-interval Rényi entropy; skippable otherwise. The real content is bookkeeping done carefully: the authors extend their twist-operator OPE machinery to CFTs with W3 and W4 symmetry and match it against the 1-loop handlebody partition function of higher spin AdS3 through order x^8 in the cross ratio. The new pieces are the spin-3 and spin-4 1-loop contributions (2.8) and (2.9) and, more usefully, the catalog of W-sector quasiprimaries with nonvanishing d_K in Appendices A and B. That table is what I would actually reach for later; the rest is a long expansion that mostly confirms what the character formulas already suggest. On the reader's worry about the classical saddle: I think it is over-weighted. In the ground state the W charges vanish, so the W fields first enter at O(c^0) — level 6 for W3 squared, level 8 for W4 squared — which is exactly where the 1-loop comparison lives. The classical part is the ordinary result and the footnote to [35] is adequate. The stress-test's explicit check of the n=2, x^6 coefficient is convincing, and the match is not an artifact of shared character formulas, because the d_K's are computed independently from conformal transformations. Soft spots, in proportion: the abstract's 'confirms' is a bit strong for one observable at one loop — 'supports' would be honest. The 2-loop and 3-loop results (3.17) and (3.18) have no gravity counterpart, which the authors say plainly; they are CFT curiosities for now. The vanishing 3-loop at n=3 is flagged as unexplained, and the speculation about m-loop vanishing for n up to m is just that. The overlap with Perlmutter [35] is disclosed in a note, which is the right handling. Nothing here is wrong. It is a grind-it-out confirmation paper, and the grind is done properly. Send it out.","headline":"A careful, grind-it-out confirmation that the 1-loop Rényi entropy of W-symmetric CFTs matches higher spin AdS3 through order x^8, with the operator tables being the durable contribution.","tokens_in":30562,"tokens_out":1487,"would_cite":true,"duration_ms":26108,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"selfhosted/deepseek-v4.1-flash","headline":"Two disjoint intervals in a W-symmetric conformal field theory carry a Rényi entropy whose gravity-side and field-theory-side computations agree exactly through order $x^8$ in the cross ratio, once massless spin-3 and spin-4 fields are tall","keywords":[],"falsifier":"Recompute the order-$x^8$ coefficient of the spin-3 one-loop contribution on the gravity side from a background in which the higher spin fields are allowed to backreact on the handlebody, and check whether the coefficient of $x^8$ in (2.8) still equals the corresponding $W_3$ term extracted from the twist-operator OPE; a mismatch in that single coefficient, or a failure of the analogous spin-4 comparison in (2.9) against (3.23), would break the claim.","tokens_in":30808,"feed_emoji":"🌀","tokens_out":1962,"duration_ms":36631,"temperature":0.7,"pith_summary":"The paper asks whether the holographic recipe for entanglement entropy survives when the boundary theory has more symmetry than ordinary conformal invariance. It studies the Rényi mutual information of two short, widely separated intervals on the plane in a two-dimensional conformal field theory with $W_3$ and $W_4$ symmetry, the kind of theory conjectured to be dual to a higher spin theory of gravity in three-dimensional anti-de Sitter space. Both sides are expanded in the small cross ratio $x$ up to order $x^8$, which is far enough to expose the leading and subleading imprints of the spin-3 and spin-4 fields. The classical (order $c$) and one-loop (order $c^0$) pieces agree exactly, and the authors also record the two-loop and three-loop pieces on the field-theory side. If the match holds, it says the Ryu-Takayanagi area law plus a one-loop sum over massless fields of every spin is the right bookkeeping for entanglement entropy in higher spin holography, not just in pure gravity.","feed_headline":"Higher-spin gravity matches CFT entropy to eighth order","feed_subtitle":"Two-interval Rényi entropy agrees between higher-spin AdS₃ and a W-symmetric conformal field theory through order x⁸.","key_machinery":"Two machines carry the argument and are matched against each other. The gravity machine is the one-loop partition function on a handlebody quotient of AdS$_3$ by a Schottky group, written as a product over primitive conjugacy classes and over massless spins $s$; because the product factorizes by spin, the graviton, spin-3 and spin-4 contributions can be computed separately and simply added. The field-theory machine is the short-interval operator product expansion of the twist operators $σ$ and $σ̃$, $σ(z)σ̃(0) = c_n Σ_K d_K Σ_{m,r} a^K_m/m! · ā^K_r/r! · z^{-2h+h_K+m} z̄^{-2h̄+h̄_K+r} ∂^m∂̄^r Φ_K(0)$, summed over all independent quasiprimary operators $Φ_K$ of the replicated theory. The coefficients $d_K$ are fixed by the conformal transformation of each quasiprimary operator, and the whole two-interval partition function collapses into a sum of hypergeometric functions $F(h_K,h_K;2h_K;x)$. The nontrivial work is enumerating the quasiprimary operators through level eight, including those that mix the stress tensor with $W$ fields, and evaluating their $d_K$.","core_discovery":"The central claim is an exact agreement, through order $x^8$, between two computations of the Rényi mutual information $I_n$ of two disjoint intervals with small cross ratio. On the gravity side the classical piece is taken over unchanged from pure AdS$_3$ gravity, and the one-loop piece is read off from a product over primitive conjugacy classes of the Schottky group, $Z^{1-!loop} = 1/|1-q_γ^m|$ extended to a product over every massless spin $s$ with $m$ running from $s$ upward; the spin-3 and spin-4 contributions are additive and are given explicitly in (2.8) and (2.9). On the field-theory side the same quantity is extracted from the operator product expansion of twist operators, now including quasiprimary operators built from the $W$ fields alongside those of the Virasoro Verma module. The result is that $I^{(2,3)}_n$ equals the ordinary Virasoro result (3.16) plus the spin-3 term (2.8), and likewise for $W_4$, with the tree-level part (3.15) untouched by the extra symmetry. The authors also obtain the $1/c$ and $1/c^2$ corrections on the field-theory side, note that they vanish at $n=2$ in line with the genus-one partition function being one-loop exact, and observe that the three-loop term appears to vanish at $n=3$ without a known reason.","pith_inferences":[],"forward_implications":[],"supporting_citations":[],"fun_headline_variants":["Higher-spin AdS₃ matches W-CFT Rényi entropy to order x⁸","W-symmetry CFT and higher-spin gravity agree to one loop","Two-interval Rényi entropy agrees through order x⁸ on both sides","Three-loop Rényi term vanishing at n=3 has no known reason","Classical Rényi piece stays unchanged under W symmetry"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The classical gravitational saddle is assumed to be unchanged by the presence of the higher spin fields, so that the leading, order-$c$ Rényi entropy is still given by the Ryu-Takayanagi area formula; the paper states this in Section 2 and points to a companion paper for a more careful justification.","fun_headline_variants_meta":{"raw":{"variants":["Higher-spin AdS₃ matches W-CFT Rényi entropy to order x⁸","W-symmetry CFT and higher-spin gravity agree to one loop","Two-interval Rényi entropy agrees through order x⁸ on both sides","Three-loop Rényi term vanishing at n=3 has no known reason","Classical Rényi piece stays unchanged under W symmetry"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2101,"prompt_tokens":1182,"completion_tokens":919,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":829}},"tokens_in":1182,"tokens_out":919,"duration_ms":16059,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":829,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-10T08:52:03.669107+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Recompute the order-$x^8$ coefficient of the spin-3 one-loop contribution on the gravity side from a background in which the higher spin fields are allowed to backreact on the handlebody, and check whether the coefficient of $x^8$ in (2.8) still equals the corresponding $W_3$ term extracted from the twist-operator OPE; a mismatch in that single coefficient, or a failure of the analogous spin-4 comparison in (2.9) against (3.23), would break the claim.","supporting_citations":[],"review_version":1}