{"id":"9979fd2e-9d90-4548-af74-cc3f64099e5d","arxiv_id":"2607.29662","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Timelike subregion volume complexity in de Sitter grows exponentially early and diverges hyperfast at a maximal duration; near the SdS black hole horizon the divergence is replaced by slower, claimed-nonlinear growth.","lead":"This paper computes the \"timelike\" holographic complexity of de Sitter space — a measure of the quantum-information cost of describing part of a spacetime — using static patch holography. It finds exponential growth for short time intervals and a divergence at a finite maximal interval, with the divergence replaced by slower growth when the stretched horizon sits near a Schwarzschild–de Sitter black hole.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"SdS Case 2 headline result is contradicted by the paper's own Appendix B: Eq. (B.6) gives C_s ~ r_min^{d-1} T (linear), not nonlinear growth.","rationale":"I read the paper in good faith. The pure de Sitter analytic core (Section 3) is coherent: the timelike entanglement entropy and timelike volume complexity are derived explicitly, and the late-time logarithmic divergence in d=3 is a concrete analytical prediction. The numerical cross-checks in Kruskal and Eddington-Finkelstein coordinates (Tables 3–6) provide genuine support for the higher-dimensional pure-dS extensions, and the early-time exponential behavior e^{2T} is a clean, parameter-free consequence of the dS3 formula. Those parts of the paper are not what I would challenge.\n\nThe load-bearing problem is the SdS Case 2 result, which is presented as the paper's novel contribution. The abstract and Section 5.2 claim that hyperfast growth is replaced by a 'nonlinear growth regime,' while Appendix B—the paper's own analytical late-time derivation for exactly that case—concludes C_s ~ r_min^{d-1} T, which is linear in T. The appendix even contains a sentence denying linear growth immediately after displaying a linear formula. This is not a subtle interpretive issue: the central claim and the supporting derivation are mutually inconsistent.\n\nIf the inconsistency is resolved in favor of Appendix B, then the headline SdS Case 2 result is false as stated and should be corrected to 'late-time linear growth with slope set by r_min^{d-1}.' If instead the numerical Fig. 17 is correct and p ≠ 1, then Appendix B's asymptotic expansion is wrong, and the paper's analytical support for its own conclusion evaporates. Either way, the current manuscript cannot be accepted without a correction. A single targeted computation—fitting the exact late-time C_s(T) from (5.36) to a power law—would settle which half of the contradiction is wrong. The reader's CONDITIONAL verdict already captures the need for such a fix, so I do not recommend changing the verdict, but I want to make explicit that the decisive fault is the Appendix B/Sec. 5.2 contradiction rather than the t=0 turning-point symmetry emphasized in the reader's weakest_assumption field.","tokens_in":31361,"tokens_out":7938,"duration_ms":76375,"concrete_test":"Evaluate C_s(T) from the exact integrals (5.36) using v'(r) from (5.27) for d=3, μ=0.1, L=3, ϵ=10^{-5} in the late-time window of Fig. 17 and fit C_s = A T^p. If p≈1, Appendix B (B.6) is confirmed and the abstract/Sec. 5.2 'nonlinear growth' claim must be retracted; if p≠1, the error lies in Appendix B, and Eq. (B.6) should be re-derived. A purely analytical version: redo the (B.4)→(B.6) step while keeping the exact r* dependence and the omitted tilde r(r*)−tilde r(r) terms; check whether C_s/T becomes r_min^{d−1} or a nontrivial function of T.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's most distinctive claim is the SdS Case 2 statement in the abstract and Section 5.2: when the stretched horizon is near the black hole, the late-time hyperfast growth is replaced by a 'nonlinear growth regime.' However, Appendix B, which is explicitly devoted to a late-time analysis of Case 2, derives in Eq. (B.6) that C_s ~ r_min^{d-1} T. This is linear in the boundary duration T. Appendix B's own concluding sentence says this 'does not show linear growth with time,' which is the opposite of its own formula. Section 5.2 and the Discussion instead describe 'nonlinear growth (mix of powers)' and a 'power law increase with the boundary subregion duration.' These statements cannot both be correct. If Eq. (B.6) is right, the abstract and Section 5.2 are wrong; if the numerical Fig. 17 really shows T^p with p ≠ 1, then the asymptotic reduction in Appendix B contains an algebraic error, most likely in the nested integral in (B.4) or in the δ → 0 limit leading to (B.6). This is an internal inconsistency, not a matter of external consensus. A second internal inconsistency affects the pure-dS result: the d=3 late-time divergence is stated as logarithmic in Eq. (1.2) and Section 3.2, but as a power law (T_max − T)^{-(d−2)} = (T_max − T)^{-1} in Eq. (1.3) and Section 4. Both cannot hold simultaneously.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies the recently proposed holographic timelike subregion volume complexity to de Sitter spacetime in the static patch holography framework. For pure dS it computes analytic dS3 expressions for the spacelike and timelike extremal surfaces, the timelike entanglement entropy, and the complexity, then extends the computation numerically to d=4,5. For Schwarzschild–de Sitter, it treats two stretched-horizon positions: near the cosmological horizon (Case 1) and near the black hole horizon (Case 2). The headline claims are that pure dS timelike complexity grows as e^{2T} for short durations and diverges hyperfastly near a maximal duration, while SdS Case 2 replaces the hyperfast growth by a 'nonlinear growth regime' for long durations.","tokens_in":31641,"tokens_out":7658,"duration_ms":72872,"significance":"If established, this would be a useful extension of timelike complexity to cosmological horizons, showing that the construction yields a finite, real, positive measure in dS and that the late-time divergence is dimension-dependent. The dS3 analytic solution, the numerical cross-checks in Kruskal and Eddington–Finkelstein coordinates (Tables 3–6, Appendix A), and the absence of fitted constants in the extremization are genuine strengths. However, the central Case 2 claim is internally inconsistent with the paper's own Appendix B, and the pure-dS late-time divergence formulae contradict each other for d=3. These issues must be resolved before the paper's main conclusions can be accepted.","major_comments":[{"comment":"The SdS Case 2 claim of 'nonlinear growth' is contradicted by Appendix B. Eq. (B.6) gives C_s ∼ r_min^{d−1} T, which is linear in T for fixed r_min, while the sentence immediately after (B.6) states that this 'does not show linear growth with time.' Section 5.2 and the Discussion instead describe 'nonlinear growth (mix of powers)' and a 'power law increase with the boundary subregion duration,' without any fitted exponent. Both statements cannot be correct. The authors must determine whether Eq. (B.6) or the numerical reading of Fig. 17 is right; if the former, the abstract and §5.2 must be revised to report linear growth; if the latter, the asymptotic reduction in (B.4)–(B.6) contains an error.","section":"Abstract; §5.2; Appendix B"},{"comment":"The pure-dS late-time divergence is stated inconsistently. Eq. (1.2) and Eq. (3.40) give C_late ∼ ln[(T_max−T)^{−1}] for d=3, while Eq. (1.3) and Eq. (4.13) give C_late ∼ (T_max−T)^{−(d−2)}, which for d=3 is (T_max−T)^{−1}, not logarithmic. Section 4 and Fig. 8 apply the power-law form to d=3,4,5. This is a direct contradiction in a central quantitative claim. If logarithmic divergence is special to d=3, the general formula must state d>3 and explain how the d=3 limit arises; if the power law is correct, the §3.2 analytic result needs re-examination.","section":"Eqs. (1.2), (1.3), (3.40), (4.13)"},{"comment":"The SdS computation relies on the assumption that the turning point of the spacelike extremal surface lies on the t=0 slice, and that the timelike connecting curve is u(r)=−r̃(r) or v(r)=r̃(r). This is asserted via 'insight based on symmetry' (§5.1) and 'we can argue from the symmetry of SdS geometry' (§5.2), rather than derived from the extremization conditions with finite boundary subregions. For pure dS this condition follows from the U↔V reflection symmetry of the maximal-radius point; for the infinitely extended SdS Penrose diagram and a finite timelike boundary interval it is not automatic. Since this assumption fixes the boundary conditions for all computed surfaces and volumes in Cases 1 and 2, a proper justification is load-bearing. The authors should either prove the t=0 turning point from the equations of motion or state the precise symmetry and why it survives the choice of a","section":"§5.1, Eq. (5.16); §5.2, Eqs. (5.29), (5.35)"},{"comment":"The evidence for 'nonlinear (polynomial)' growth is not quantified. Fig. 17 plots ln C_s versus T, not versus ln T, and shows only two mass values with no error bars, no fitted exponent, and no alternative linear fit. Given the conflict with Appendix B, a quantitative late-time analysis is necessary: for example, extract d ln C_s/d ln T at large T, or directly test C_s = a T + b and compare with a power law. The present plot does not by itself support a 'mix of powers' claim.","section":"§5.2, Fig. 17"}],"minor_comments":[{"comment":"The expression contains 'ln ln' where a single logarithm is clearly intended; the formula should read C_s = (4c/3) Tϵ + (4c/3) ln(...).","section":"Eq. (3.36)"},{"comment":"The caption mentions 'AdS radius L' in a de Sitter context; this should be 'de Sitter radius L'.","section":"Fig. 13 caption"},{"comment":"The exponent is written as '(d−2)−#g(μ)' with the symbol '#' unexplained and the function g(μ) not defined. Please define g and specify the range of T_max−T over which this fit is valid.","section":"Eq. (5.21)"},{"comment":"The notation V_{S^{d−1}} is used without definition; it should be identified as the volume of the unit (d−1)-sphere.","section":"§2"},{"comment":"Several references use incomplete entries, e.g., [1] is listed with only an arXiv number and no year, and [82] repeats the arXiv identifier. Please normalize the bibliography.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The inconsistency between Appendix B and Section 5.2 is severe and directly affects the paper's most distinctive claim. The authors also need to address the d=3 power-law/logarithmic conflict. If the t=0 turning-point assumption cannot be justified for finite SdS subregions, the SdS sections would require substantial reworking. These are fixable in principle, but only with a serious revision rather than cosmetic changes."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new and useful part is the computation of timelike subregion complexity in pure de Sitter static patch holography. The d=3 analytic formulas for complexity and timelike entanglement entropy are concrete, and the numerical checks in two coordinate systems give real support. The result that this quantity is real, finite, and grows as e^{2T} at early times and diverges at T_max is a nice data point for the timelike-complexity proposal. If you work on dS holography or entanglement probes, the pure dS sections are worth your time.\n\nThe soft spot is exactly where the stress-test note lands. The paper's most distinctive claim is that in SdS with the stretched horizon near the black hole horizon (Case 2), hyperfast growth is replaced by a nonlinear growth regime. But Appendix B, which is supposed to derive that regime analytically, ends with Eq. (B.6): C_s ~ r_min^{d-1} T. That is linear in T. The very next sentence says this \"does not show linear growth with time,\" which is the opposite of the formula. Section 5.2 and the abstract call it \"nonlinear growth (mix of powers).\" These cannot all be correct. Either the appendix has an algebraic slip and the numerics are right, or the numerics are misleading and the appendix is right. A referee needs to see this resolved, not papered over.\n\nThere is a smaller inconsistency in the pure dS result: the d=3 late-time divergence is stated as logarithmic in Eq. (1.2) and as a power law (T_max - T)^{-(d-2)} in Eq. (1.3). The text says the log is a \"dimensional accident,\" but Eq. (1.3) as written applies to d=3 too. That is a presentation bug.\n\nTwo other items are less severe but worth noting. The turning-point-at-t=0 assumption is used throughout the SdS analysis without an independent derivation; it is plausible by symmetry, but the paper would be stronger if it showed why it holds for finite boundary subregions. Also, the numerical scaling exponents in the Case 2 plots are quoted without error bars, so the claim that the late-time growth is a power law with a specific exponent is not as firm as the text suggests.\n\nBottom line: the pure dS part is a solid contribution and deserves to be in the literature. The SdS Case 2 story is internally inconsistent and needs a careful revision. I would send this to peer review, but the referee should insist on reconciling Appendix B with the abstract and Section 5.2, and on clarifying the d=3 divergence statement.","headline":"The pure dS calculation is solid and worth reading, but the SdS Case 2 headline is contradicted by the paper's own Appendix B, so the abstract and Section 5.2 need correction before this can be trusted.","tokens_in":32285,"tokens_out":2160,"would_cite":true,"duration_ms":22580,"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":"Timelike subregion complexity in de Sitter is real and finite, grows exponentially at early durations, and diverges hyperfast near a maximal duration unless a black hole horizon intervenes.","keywords":["timelike subregion complexity","holographic complexity","de Sitter static patch","volume complexity","Schwarzschild-de Sitter","timelike entanglement entropy","extremal surfaces","stretched horizon"],"falsifier":"For the Schwarzschild–de Sitter Case-2 geometry, numerically integrate the extremal-surface equations without imposing U=V (or v(r*) = r̃(r*)) at the turning point; if the resulting late-time complexity growth differs from the claimed power-law/linear regime, the symmetry assumption is the point of failure. Separately, a late-time run with epsilon and T-max approached to high precision can settle whether C_s grows like r_min^{d-1} T or like a genuinely nonlinear power of T, since the paper currently asserts both.","tokens_in":31147,"feed_emoji":"⏳","tokens_out":7182,"duration_ms":64512,"temperature":0.7,"pith_summary":"The paper extends the holographic 'complexity equals volume' idea to timelike subregions of the stretched horizon in de Sitter static-patch holography. It argues that this timelike subregion complexity is a genuine observable: real, positive, and free of ultraviolet divergences, unlike timelike entanglement entropy, which picks up an imaginary part. In pure de Sitter it grows as e^{2T} for short subregion durations and diverges as (T_max - T)^{-(d-2)} (logarithmically in d=3) near a maximal duration T_max set by the stretched-horizon cutoff. For Schwarzschild–de Sitter with the stretched horizon near the cosmological horizon the same qualitative behavior persists with mass-dependent exponents; with the stretched horizon near the black hole horizon, the finite-time blow-up disappears and late-time growth becomes a power law. If correct, this gives a boundary-anchored, causally well-defined measure of complexity that can probe the black hole interior.","feed_headline":"Black hole horizon tames de Sitter complexity blow-up","feed_subtitle":"Real and finite timelike subregion complexity: exponential early growth, then a blow-up that a black-hole horizon slows.","key_machinery":"The construction has three parts: (i) the stretched-horizon prescription, which places the holographic boundary just inside the cosmological horizon and regulates the maximum boundary duration as T_max = (1/2) ln((2-epsilon)/epsilon); (ii) the timelike-subregion complexity prescription, in which two spacelike extremal surfaces anchored at the past and future endpoints and one timelike extremal surface joining their turning points bound a bulk volume; and (iii) the reflection-symmetry assumption that the turning points lie on the t=0 slice (U=V in Kruskal coordinates), which fixes all boundary conditions. In pure dS_3 the extremal surfaces are known analytically; for higher dimensions and for","core_discovery":"On the paper's own terms, the discovery is that timelike-subregion volume complexity is computable and well-defined in de Sitter static-patch holography. For a boundary subregion of duration 2T on the stretched horizon, the complexity is the bulk volume bounded by two spacelike extremal surfaces anchored at the subregion's endpoints and a timelike extremal surface joining their turning points. In pure dS_3 this volume is analytic, real and positive; in higher dimensions it is numerical, checked in Kruskal and Eddington–Finkelstein coordinates. The observable reproduces the early exponential and late hyperfast growth of spacelike volume complexity, with the divergence exponent lowered by one,","pith_inferences":["A testable extension: compute timelike subregion complexity for rotating or charged de Sitter black holes; if the Case-2 softening persists, the black-hole-horizon effect is generic rather than a quirk of Schwarzschild–de Sitter.","The e^{2T} early growth matches the behavior expected of Krylov-type operator growth, so timelike subregion complexity may be a more direct holographic proxy for operator growth in the static patch than full volume complexity.","The paper's Case-2 late-time result is stated in two ways: the abstract and discussion describe nonlinear growth, while the appendix derives C_s ~ r_min^{d-1} T, linear in T. Since these are not the same statement, a high-precision late-time numerical check is needed to decide which behavior the extremal-surface construction actually produces.","If the turning-point-at-t=0 symmetry is relaxed, the entire Case-2 late-time scaling would shift; testing this assumption in the numerical integration would reveal whether the claimed black-hole-horizon softening is robust."],"forward_implications":["If the proposal is correct, timelike subregion complexity is a real, positive, finite holographic quantity in de Sitter, with no ultraviolet divergence, in contrast to timelike entanglement entropy which is complex.","In pure de Sitter, the observable grows as e^{2T} at short durations and diverges as (T_max - T)^{-(d-2)} near the maximal duration, with T_max identical to the critical time of full-volume complexity.","For Schwarzschild–de Sitter with the stretched horizon near the cosmological horizon, the same exponential-then-hyperfast pattern holds, with the black-hole mass increasing T_max and modifying the divergence exponent.","With the stretched horizon near the black hole horizon, timelike complexity does not diverge at finite time; its long-duration growth is a power law (per the paper's abstract and discussion) or linear with coefficient r_min^{d-1} (per the appendix), indicating the black hole interior rather than the expanding cosmological region.","The growth-rate pattern — e^{T} for full volume complexity and e^{2T} for timelike subregion complexity — suggests the rate tracks the codimension of the boundary anchoring surface, giving a new organizational rule for holographic complexity."],"fun_headline_variants":["Black hole slows de Sitter complexity blow-up","Timelike complexity in de Sitter: hyperfast growth tamed by black hole","De Sitter timelike complexity: real, finite, hyperfast growth","Holographic timelike complexity for de Sitter computed","Black hole horizon reins in de Sitter complexity"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire calculation assumes that the turning point of every spacelike extremal surface lies exactly on the t=0 slice (U=V), argued from the left-right symmetry of the extended Penrose diagram but not independently derived for finite boundary subregions; if the turning point is off t=0, every boundary condition, volume, and late-time scaling in the paper changes.","fun_headline_variants_meta":{"raw":{"variants":["Black hole slows de Sitter complexity blow-up","Timelike complexity in de Sitter: hyperfast growth tamed by black hole","De Sitter timelike complexity: real, finite, hyperfast growth","Holographic timelike complexity for de Sitter computed","Black hole horizon reins in de Sitter complexity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00054,"raw_usage":{"total_tokens":2409,"prompt_tokens":709,"completion_tokens":1700,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":453,"completion_tokens_details":{"reasoning_tokens":1615}},"tokens_in":453,"tokens_out":1700,"duration_ms":11848,"temperature":1.0,"reasoning_tokens":1615,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T02:20:41.166577+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For the Schwarzschild–de Sitter Case-2 geometry, numerically integrate the extremal-surface equations without imposing U=V (or v(r*) = r̃(r*)) at the turning point; if the resulting late-time complexity growth differs from the claimed power-law/linear regime, the symmetry assumption is the point of failure. Separately, a late-time run with epsilon and T-max approached to high precision can settle whether C_s grows like r_min^{d-1} T or like a genuinely nonlinear power of T, since the paper currently asserts both.","supporting_citations":[],"review_version":1}