{"id":"7500ea75-de4b-452d-817b-a6b8de463a7b","arxiv_id":"2505.01247","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"For strip-shaped CFT_d systems, bath-pair entropy peaks at a generalized golden-ratio critical size, then falls through an exactly resummed island entropy term, with far-separated mutual information decaying as 1 over distance to the power d.","lead":"This paper derives formulas for how the entanglement entropy of two identical bath regions in a conformal field theory changes as their size grows while the total system stays fixed. The result gives closed-form island corrections and critical size ratios that generalize the golden ratio, aimed at modeling Page-curve behavior in holographic systems.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exact-island claim for d>2 collapses if the assumed mutual-information identity (26) fails; compactification to d=2 does not justify it, and the RT phase structure for separated strips in d>2 makes it doubtful.","rationale":"The reader's weakest_assumption already identifies eq. (26) as the load-bearing premise, and I agree. My pass adds two observations: the compactification justification is not valid, because a d-dimensional strip entropy is area-law rather than a 2d logarithmic form; and eq. (26) predicts nonzero mutual information for arbitrarily separated strips, contrary to the standard RT phase structure for disjoint regions in d>2. I do not claim the paper is internally inconsistent: the algebra from (26) to (32) is correct, and the paper explicitly flags the assumption. The right disposition is therefore the reader's CONDITIONAL verdict, unchanged, with the proposed numerical RT check as the concrete condition. If the check rules out eq. (26), the verdict should move to REJECT; if eq. (26) is unexpectedly confirmed, the central construction is correspondingly supported.","tokens_in":15894,"tokens_out":19894,"duration_ms":208730,"concrete_test":"Check eq. (26) directly in AdS4: numerically solve the RT minimal-surface equations for two disjoint strips of width b=1 separated by 2a=10 (in units of L), considering both the disconnected configuration (area 2S(b)) and connected homologous candidates. If the minimal surface is disconnected, then I(B1:B2)=0, contradicting the positive law in (27); repeat at b/a = 0.1, 0.2, 0.5 to locate the phase transition. An independent analytic check would be to derive the two-strip mutual information from the known holographic phase structure and compare it with the compactification-based formula.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equations (31)-(32) and (36)-(37) are algebraic consequences of S[B]=2S(b)-I(B:B) and the assumed identity I(B:B)=2S(2a+b)-S(l)-S(2a), eq. (26). Everything in the exact-island section, including (33), (45), and the post-critical identity (52), therefore inherits any error in (26). The paper's footnote defends (26) by claiming strip systems reduce to d=2 by compactification, but eq. (24) has an area-law divergence V/epsilon^{d-2} and a power-law finite term, which a compactified 2d CFT with logarithmic entropy does not reproduce; the reduction would introduce KK towers, not the same theory. Moreover, in AdS_{d+1} the entropy of two disjoint strips is the minimum over disconnected and connected RT surfaces. For b << 2a, the disconnected configuration with area 2S(b) is the global minimum, so the leading mutual information is zero, while eq. (26) gives a positive O(b^2/(2a)^d) value (eq. (27)). Thus eq. (26) is not merely unproved; it is inconsistent with the standard RT phase structure in d>2. If it fails, the exact island resummation, the island location (37), and the never-vanishing-MI law are unsupported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper considers a symmetric arrangement of two identical strip-shaped bath systems B on either side of a central system A in a CFT_d, with fixed total width l = 2a + 2b, and studies the bath entanglement entropy S[B] as a function of b. It claims that for d > 2 the bath entropy reaches a maximum at a 'critical ratio' satisfying a generalized Fibonacci-type equation (28), and that beyond this critical point the entropy can be written exactly as S[B] = S(l) + S(2a) - S_island(a,b), where S_island is a positive definite closed-form 'exact island' contribution (32) that resums all subleading 'iceberg' terms. It further claims an exact island location z = \\bar b (37), a mutual-information law I(B:B) \\propto b^2/D^d that never vanishes for finite systems (27), a post-critical identity S[B]-S[A] = S_l - S_island (52), and a discrete entropy spectrum when system-A approaches the Kaluza-Klein scale (Section 5). The central technical input is the mutual-information identity I(B:B) = 2S(2a+b) - S(l) - S(2a) in Eq. (26), which the paper assumes in footnote 7 for strip systems.","tokens_in":16320,"tokens_out":3946,"duration_ms":42107,"significance":"If Eq. (26) were correct, the paper would provide an unusually simple and explicit 'exact island' resummation in higher-dimensional CFTs, together with a family of generalized golden-ratio conditions and a concrete prediction for the mutual information of separated strips. The manuscript is transparent about its main assumption (footnote 7), states all algebraic steps explicitly, and supports its claims with detailed formulas and plots; the derivations from (26) to (32) and (52) are elementary and checkable. However, the significance is entirely conditional on (26), and the paper itself concedes that this identity is assumed rather than derived for d > 2. Because the subsequent exact-island results are algebraic consequences of (26), the central claim stands or falls on that identity, which the paper does not justify with a valid argument.","major_comments":[{"comment":"Equation (26), I(B:B) = 2S(2a+b) - S(l) - S(2a), is load-bearing: the critical equation (28), the exact island formula (32), the island location (37), and the identity (52) all follow algebraically from it once S(x) is given by the strip entropy (24). The manuscript's only defense is the statement in footnote 7 that the expression 'would be true at least for strip systems cases, as these can be reduced to 2-dim by compactification.' That reduction is not demonstrated and is doubtful: the strip entropy (24) has an area-law divergence V/\\epsilon^{d-2} and a power-law finite term, whereas a compactified 2d CFT has logarithmic entropy and would produce Kaluza-Klein towers rather than the same theory. Since the paper presents no derivation of (26), all of the exact-island claims inherit a substantial unsupported premise.","section":"Section 3, Eq. (26) and footnote 7"},{"comment":"Equation (26) appears to be inconsistent with the standard Ryu-Takayanagi phase structure for two disjoint strips in AdS_{d+1} with d > 2. For b \\ll a, the disconnected RT surfaces anchored on the two individual strips give S[B] \\approx 2S(b), so the mutual information is exponentially small in the holographic large-N limit and zero at the classical extremal-surface level, because connected surfaces have larger area and are not the minimum. In contrast, Eq. (26) combined with Eq. (24) yields the positive leading mutual information I(B:B) \\simeq I_0 b^2/(2a)^d stated in Eq. (27). The paper does not explain how connected RT surfaces can dominate over disconnected ones when the separation 2a is much larger than the strip width b, and no minimality argument is given for (26). This is not merely a missing proof but a concrete conflict with the RT prescription on which the paper relies.","section":"Section 3, Eqs. (24)-(27) versus standard RT phase structure"},{"comment":"The post-critical identity S[B] - S[A] = S_l - S_island is presented as an important result, but it is tautological. Equation (31) already defines S[B] as S(l) + S(2a) - S_island(a,b), and since S[A] = S(2a) is used throughout, Eq. (52) is obtained by simply subtracting S(2a) from both sides of (31). It therefore contains no independent physical content beyond the definition of S_island as the residual required to make (31) hold. The statement that 'changes in (S[B]-S[A]) would have to be compensated precisely by the islands entropy only' is a restatement of (31), not a new constraint. A genuine physical prediction would require an independent derivation of S_island from gravity or from a microscopic calculation, which is not provided.","section":"Section 6, Eq. (52)"},{"comment":"The gravitational interpretation of the exact island is selected by parameter choices rather than derived. The dictionary relating the CFT size a to the dilaton boundary value Φ_0 and the compactification radius R, e.g. a = πRΦ_0/(2^2 L b_0^2) in (38) and a = πRΦ_0/(2^2 L b_0^3) in (42), is introduced with coefficients chosen to match the algebraic expression (36). The appendix explicitly states that 'one may have to tune final relations with factors of 2 and π.' Because Φ_0 and R are free parameters and the numerical factors are adjustable, the claim that the island term (36) is the gravitational entropy of a boundary at z = \\bar b is an interpretation imposed by matching, not a prediction derived from the setup. This weakens the claim that the island location (37) and (39) have independent geometric significance.","section":"Section 4 and Appendix A, Eqs. (38), (42), (54)"}],"minor_comments":[{"comment":"The text contains numerous typographical spacing errors, including 'CF Td', 'syst em-A', and 'e-Print:' in references; these should be corrected.","section":"Throughout"},{"comment":"The figures lack axis labels and units in most cases; for example, Figure 7 shows 'b0' on the horizontal axis but the text uses b as the variable, and the vertical axis is not described as S_bath until some later figures. Please clarify the plotted quantities and parameters.","section":"Figures 3-10"},{"comment":"The recurrence C_n = 2C_{n-1} + 1 with C_{-n} = 0 is confusingly stated; the range of n and the definition of C_0 should be made explicit, and the claim that 'C_0 = 1, C_{-n}=0' should be reconciled with the recurrence for n=0.","section":"Eq. (40)"},{"comment":"The notation for the multi-strip entropy is inconsistent: Eq. (47) writes S[A]_{2n-strips}, while Eq. (50) labels the same quantity S[A]_{2n-strips} but the text refers to 'assembly of 2n narrow strips' and separately to 'n strips'. Please clarify whether n denotes the number of strips or half-strips.","section":"Section 5"},{"comment":"The text says the inequality S(l) - |S[B]-S[A]| \\geq 0 'turns into exact equality given by (31)', but (31) is not an equality involving S[A]; the logical connection between the inequality and the exact island formula should be spelled out.","section":"Eq. (33)"},{"comment":"Reference [20] appears not to be cited in the text; please check the citation list for completeness.","section":"References"}],"recommendation":"reject","confidential_remarks":"The manuscript is candid about its main assumption in footnote 7, but that assumption is not merely unproved: it conflicts with the standard RT phase structure for separated strips in d > 2, as shown by the disconnected-surface minimum for b \\ll a. Since the exact island formula and the critical ratio equations are direct algebraic consequences of Eq. (26), the central result is unsupported in a way that cannot be repaired by local edits. The paper would need a new derivation or a different starting identity to make the exact-island claim viable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: the central claim of an 'exact island' for CFT_d with d>2 is built entirely on the assumed identity for the mutual information of two disjoint strips, eq (26). The algebra from there is clean, but that identity is not just unproved—it contradicts the standard Ryu-Takayanagi phase structure for separated strips in d>2. So the exact island formula (32), the island location (37)/(43), and the never-vanishing mutual-information law should be treated with real skepticism.\n\nThe paper does some things well. It generalizes the author's earlier CFT3 construction to general d, and the critical-ratio equations (28)-(29) are a nice algebraic pattern. The text is also honest: the key assumption is flagged in a footnote, not hidden. The resummation of 'iceberg' terms into a single island term is an interesting formal exercise, and the plots are consistent with the formulas as written.\n\nThe soft spot is load-bearing. In d>2, the entanglement entropy of two small strips separated by a large distance is governed by the disconnected RT surface, so the mutual information vanishes in the classical limit. Equation (26) instead gives a positive O(b^2/(2a)^d). That is not a minor correction; it flips the sign of the very term that drives the 'exact island.' The footnote's compactification defense does not save it: a compactified d-dimensional CFT has Kaluza-Klein towers and power-law area divergences, not the logarithmic structure of a genuine 2d CFT. So the assumption is not merely unverified; it is in tension with the standard minimality rule for RT surfaces. Also, eq (52) is essentially a tautology—it restates the defining decomposition (31) with S[A]=S(2a). Minor, but worth knowing.\n\nWho is this for? Someone curious about how far a formal ansatz can go under a strong assumption, or someone tracking the author's previous work on islands and icebergs. The critical-ratio part could be of independent interest, but the island results should not be cited as established until eq (26) is derived or checked independently.\n\nI would not desk-reject this. The issue is specific, checkable, and a competent referee could quickly settle whether (26) holds for strip geometries in d>2. If it fails, the island sections should be rewritten as a heuristic model rather than a derivation. If it holds, the exact-island formula becomes a genuine result. That is exactly what peer review should sort out.","headline":"The paper's 'exact island' formula for d>2 rests on an unproved mutual-information identity that likely fails in standard RT setups, so the island results are conditional at best; the critical-ratio algebra is neat but secondary.","tokens_in":16758,"tokens_out":4000,"would_cite":false,"duration_ms":41761,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"For strip-shaped CFT_d systems with d>2, the paper derives an exact island entropy that resums all subleading 'iceberg' corrections, so the post-critical bath entropy obeys $S[B]=S(l)+S(2a)-S_{\\mathrm{island}}$.","keywords":["entanglement entropy","holographic CFT","entanglement islands","iceberg entropy","mutual information","Page curve","golden ratio","CFT_d strip systems"],"falsifier":"Compute the full bath-pair entropy $S[B]$ holographically for a CFT_3 strip with fixed total length $l$ and all bath sizes $b$, including connected RT surfaces; check whether the maximum occurs at the predicted $x_c\\simeq0.88$ and whether $S[B]=S(l)+S(2a)-S_{\\mathrm{island}}(a,b)$ holds past the maximum. A mismatch in either check, or the same test in CFT_4 with $x_c\\simeq0.96$, would falsify the exact-island claim.","tokens_in":15642,"feed_emoji":"🧊","tokens_out":10980,"duration_ms":106214,"temperature":0.7,"pith_summary":"This paper studies two identical CFT_d strips placed symmetrically around a central system A, with the total size fixed while the two bath halves B grow. For strip-shaped systems it claims the bath-pair entropy reaches a maximum when the size ratio satisfies a Fibonacci-type critical condition, and then falls in the post-critical regime. The central result is that for d>2 all subleading 'iceberg' contributions can be resummed into one exact, positive-definite island entropy, giving $S[B]=S(l)+S(2a)-S_{\\mathrm{island}}(a,b)$. If this is right, the Page-curve-like decrease of bath entropy becomes a closed-form statement rather than a perturbative series, and the identity $S[B]-S[A]=S_l-S_{\\mathrm{island}}$ holds at every post-critical point. A reader should care because one geometric object replaces an otherwise uncontrollable infinite sum of corrections.","feed_headline":"One exact formula captures falling bath entropy","feed_subtitle":"In strip-shaped CFT_d systems, all iceberg corrections resum into a single island entropy, and critical ratios generalize the golden ratio.","key_machinery":"The mechanism is the decomposition $S[B]=2S(b)-I(B:B)$, paired with the assumed mutual-information identity $I(B:B)=2S(2a+b)-S(l)-S(2a)$, where $S(x)$ is the holographic strip entropy of eq. (24). In the post-critical regime this identity is rearranged into $S[B]=S(l)+S(2a)-S_{\\mathrm{island}}$. The resummation itself is carried by the finite geometric-series identity $1-(1+2s)^{-n}=2s\\sum_{\\alpha=1}^{n}(1+2s)^{-\\alpha}$, which for $n=d-2$ converts the infinite iceberg expansion into one term proportional to $b^{2-d}$ and fixes the island boundary at $z=\\bar b$. This machinery is what turns a perturbative expansion into an exact statement.","core_discovery":"On the paper's own terms, the claim is that for CFT_d strip systems with d>2, beyond the critical bath size the entanglement entropy of the symmetric two-part bath is exactly $$S[B]=S(l)+S(2a)-S_{\\mathrm{island}}(a,b),$$ with $S_{\\mathrm{island}} = \\frac{L^{d-1}V^{(d-2)}}{(d-2)G_{d+1}}\\,2^{d-2} b_0^{d-1} b^{2-d}(1-(1+2s)^{2-d})$ and $s=a/b$. This island term is positive, UV-finite, and encodes the resummed effect of all iceberg configurations, with its boundary located at the exact holographic coordinate $z=\\bar b$ fixed by $1/\\bar b^{d-1}=b^{1-d}\\sum_{\\alpha=1}^{d-2}(1+2s)^{-\\alpha}$. The paper further establishes the post-critical identity $S[B]-S[A]=S_l-S_{\\mathrm{island}}$ and the critical-ratio equation $1/x^{d-1}-1/(1+x)^{d-1}=1$, whose $d=2$ solution is the golden ratio and whose higher-dimensional solutions approach $1$.","pith_inferences":["My inference: the geometric-series resummation is special to strip geometries; for spherical or disk-shaped subsystems the bath is connected and the same closed-form island location should not be expected.","My inference: a direct numerical test of the assumed identity (26) in free CFT_3 or CFT_4 would be decisive; the first discrepancy, if any, should appear at subleading order in the island expansion.","My inference: the critical equation has the form 'product equals difference' for $x^{d-1}$ and $(1+x)^{d-1}$; one could look for these algebraic numbers as extrema of other information measures, such as tripartite information, in the same geometry.","My inference: the claim that $I(B:B)$ never vanishes implies a lower bound on residual correlation at any finite separation; this could be tested in a lattice simulation by measuring mutual information as a function of $D$ at fixed $b$."],"forward_implications":["The falling branch of the bath-entropy Page curve becomes an exact algebraic statement: once $b$ exceeds the critical value, $S[B]$ is fully determined by $S(l)+S(2a)-S_{\\mathrm{island}}$.","The critical size ratio in $d$ dimensions satisfies $1/x^{d-1}-1/(1+x)^{d-1}=1$; for $d=2$ this is the golden ratio, for $d=3$ about $0.88$, for $d=4$ about $0.96$, approaching $1$ as $d$ grows.","For far-separated bath strips, mutual information obeys $I(B_1:B_2)\\propto b_1 b_2/D^d$ and is strictly positive for any finite separation.","After the crossover, changes in the entropy difference $S[B]-S[A]$ are compensated one-to-one by the exact island entropy: $S[B]-S[A]=S_l-S_{\\mathrm{island}}$.","When the central system shrinks to the Kaluza-Klein scale, bath entropy becomes discrete, with jumps indexed by the number $n$ of strips exchanged."],"supporting_citations":[{"why":"Supplies the exact two-interval entanglement entropy and mutual-information formula in CFT2, the template for the bath decomposition used throughout.","marker":"[17]"},{"why":"Provides the holographic strip entanglement-entropy formula (24) that defines every S(x) in the higher-dimensional argument.","marker":"[21]"},{"why":"Establishes the extremal-surface holographic prescription by which all strip entropies are computed.","marker":"[18, 19]"},{"why":"Defines the island and generalized-entropy setup that this paper extends by resumming iceberg corrections.","marker":"[3]"},{"why":"Introduces the island and iceberg contributions in CFT2 and CFT_d whose resummation yields the exact island formula.","marker":"[15, 16]"},{"why":"Gives the Araki-Lieb entropy inequality that the paper's finite-system inequality (33) parallels.","marker":"[22]"}],"fun_headline_variants":["Exact island formula sums all iceberg corrections","Beyond criticality, bath entropy falls exactly in CFT_d","Golden ratio yields to higher critical ratios in CFT_d","One exact island term replaces infinite iceberg series"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the mutual-information identity $I(B:B)=2S(2a+b)-S(l)-S(2a)$, exact for two intervals in two-dimensional CFT, continues to hold for CFT_d strip systems with $d>2$; the paper states this 'would be true at least for strip systems cases', but if it fails, the critical equation, the exact island formula, and the entropy-difference identity all collapse.","fun_headline_variants_meta":{"raw":{"variants":["Exact island formula sums all iceberg corrections","Beyond criticality, bath entropy falls exactly in CFT_d","Golden ratio yields to higher critical ratios in CFT_d","One exact island term replaces infinite iceberg series"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000316,"raw_usage":{"total_tokens":1839,"prompt_tokens":1043,"completion_tokens":796,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":659,"completion_tokens_details":{"reasoning_tokens":743}},"tokens_in":659,"tokens_out":796,"duration_ms":8494,"temperature":1.0,"reasoning_tokens":743,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:23:55.278495+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full bath-pair entropy $S[B]$ holographically for a CFT_3 strip with fixed total length $l$ and all bath sizes $b$, including connected RT surfaces; check whether the maximum occurs at the predicted $x_c\\simeq0.88$ and whether $S[B]=S(l)+S(2a)-S_{\\mathrm{island}}(a,b)$ holds past the maximum. A mismatch in either check, or the same test in CFT_4 with $x_c\\simeq0.96$, would falsify the exact-island claim.","supporting_citations":[{"cited_title":"Entanglement Inequalities","cited_arxiv_id":null,"evidence_quote":"Gives the Araki-Lieb entropy inequality that the paper's finite-system inequality (33) parallels."}],"review_version":1}