{"id":"7d50ecca-65f2-4ea6-9f7e-d3b06c35c921","arxiv_id":"2608.12472","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For fixed-energy PET states, the relative Wigner negativity of the boundary subregion is exp[(A_out - A_min)/(8G_N)], giving an exponential enhancement of stabilizer complexity when a python's lunch is present.","lead":"This paper calculates how much 'magic' (a non-stabilizer resource) is present in the reduced state of one side of a partially entangled thermal state in holographic quantum gravity. It finds that the magic grows exponentially with the difference in area between the outer and minimal extremal surfaces, connecting the python's lunch geometry to the hardness of classically simulating the boundary state.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Assumption 1 is tested only at the level of diagonal marginals; the full Haar randomness of U is load-bearing for eq. (46) and remains unverified.","rationale":"Both the reader and I locate the decisive risk in Assumption 1. The paper's internal derivation is coherent: given a Haar-random U and Gaussian random R_ij, eqs. (38) and (44) are standard, and the integral-representation check in Appendix A is a genuine independent confirmation of the replica-trick formula. The paper is also honest about its assumptions and about the restriction to fixed-energy boundary conditions. My concern is narrower but load-bearing: every step that removes the computational-basis dependence is an average over U, and the only numerical evidence checks a single marginal of U. Wigner negativity is a global, basis-dependent observable; two states with identical S2 can have vastly different magic in a fixed basis, so the claimed exponential enhancement is a statement about Haar-typicality of the microcanonical subspace, not about S2 alone. The SYK test of the diagonal projection probability does not constrain the off-diagonal phases and higher moments that enter the phase-point operator contractions. In addition, the SYK model is time-reversal-symmetric (real eigenvector matrix), whereas the assumption as used is complex unitary, so the evidence is not even for the same ensemble. These considerations keep the paper at moderate confidence and conditional acceptance; they do not invalidate the calculation as a conditional statement. I recommend no change to the reader's verdict.","tokens_in":16707,"tokens_out":22077,"duration_ms":221462,"concrete_test":"Compute four-point correlations of the exact microcanonical eigenvector matrix in the SYK model of Appendix B (e.g., <U_{q,i}U*_{q,i}U_{q',j}U*_{q',j}> for distinct q,q') and compare with the Weingarten/Haar expectation of eq. (29). Then compute the actual Wigner negativity (or a qubit-compatible magic monotone) of the right-reduced fixed-energy PET state for many disorder realizations and window sizes, and compare mean and variance with eq. (38) and the claimed sigma^2 ~ 1/(D e^{S2}). Deviation beyond the estimated fluctuations would show that the Beta-marginal evidence of Fig. 2 is insufficient for the full Haar assumption.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Assumption 1 in Sec. III states that U_{q,i} = <q|E_i> is well approximated by a Haar-random unitary. This is the load-bearing premise of the entire calculation: eq. (38) follows from averaging |W(u)| over U and using the Weingarten contractions of eq. (29), which require the joint distribution of all entries of U, not just their marginal moduli. The final exponential in eq. (46) inherits this premise, since the cancellation of D leaves exp[(1/2)(S2(rho0)-S2(rho))] only after the Haar-average computation. The numerical evidence in Appendix B, however, tests only x = <q|P|q> = sum_i |U_{q,i}|^2 against a Beta distribution (B3). That statistic depends solely on the row norms of U; it is insensitive to the phases and higher-order correlations among entries that enter the phase-point operator contractions in eqs. (33)-(37). A matrix can reproduce the Beta marginal exactly while being far from Haar, and a state with the same S2 can have zero Wigner negativity if it is stabilizer in the chosen basis. Thus the leap from 'chaotic eigenvectors look random' to 'the full D x k isometry is Haar' is unverified, and this is the condition on which the claimed basis-independence and the exponentially enhanced stabilizer complexity rest. A further hidden input is that in the actual CFT, rho_R and U both derive from the same Hamiltonian, whereas the Haar average treats rho_R as fixed and independent; the paper does not address their joint statistics. None of this makes the derivation internally inconsistent, but it means the headline result is conditional on a much stronger random-matrix assumption than the presented SYK test supports.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies the stabilizer complexity—quantified by Wigner negativity—of the reduced density matrix obtained from one side of a fixed-energy partially entangled thermal (PET) state in a holographic CFT. Under two explicit randomness assumptions (Assumption 1: the energy eigenbasis is Haar-random relative to a computational basis; Assumption 2: the heavy-operator matrix elements are i.i.d. Gaussian), the authors derive the relative Wigner negativity N(ρ_R|ρ_R^(0)) ~ exp[(A_out − A_min)/(8G_N)] (Eq. (46)). The derivation reproduces the Haar-random-state formula N ~ sqrt(D/e^{S_2}) using a replica trick, gives an alternative integral-representation derivation in Appendix A, computes the second Rényi entropy of the PET state by random tensor network arguments, and presents SYK numerics in Appendix B testing the diagonal projection statistic ⟨q|P|q⟩. The main result is interpreted as exponential enhancement of stabilizer complexity exactly when the bulk entanglement wedge contains a python's lunch.","tokens_in":16990,"tokens_out":9386,"duration_ms":88093,"significance":"If the result holds, it provides a concrete boundary quantity—the relative Wigner negativity—that is exponentially large precisely in the presence of a python's lunch, and it is claimed to be independent of the computational basis within the random-matrix window. The paper is clearly written and does useful work in reducing the negativity computation to a second-Rényi computation, with a second derivation in Appendix A and a numerical check in Appendix B. The main caveat is that the central formula is conditional on two strong randomness assumptions; the evidence for the more important one, Assumption 1, is only partial. The result is significant as a conjectural bridge between quantum magic and holographic complexity, but it is not yet a proof.","major_comments":[{"comment":"The numerical evidence for Assumption 1 tests only the distribution of x = ⟨q|P|q⟩ = Σ_i |U_{q,i}|^2 (Eq. (B2)), which depends only on the row norms of U. The derivation of Eq. (38), in contrast, uses the full Weingarten formula (29) and the contraction rules (33)–(37), which require the joint Haar distribution of all entries of U, including phases and higher-order correlations. A matrix can reproduce the Beta row-norm marginals exactly without being Haar-distributed, and a state with the same second Rényi entropy can have zero Wigner negativity if it is a stabilizer state in the chosen basis. The paper should either test higher-order Haar statistics (for example off-diagonal entry distributions, four-point correlators, or unitary k-design measures) or explicitly state that Assumption 1 remains unverified beyond the diagonal marginal, and correspondingly weaken the basis-independence claim.","section":"Section III (Assumption 1) and Appendix B"},{"comment":"The Haar average is performed with ρ_R held fixed, as in Eqs. (26) and (38), but in the actual CFT ρ_R and U both originate from the same Hamiltonian: U is the energy-eigenvector matrix of H_R, while ρ_R is constructed from matrix elements of the heavy operator in that same energy eigenbasis (Eqs. (10)–(11)). The paper does not address the joint statistics of U and ρ_R. If the eigenvector matrix and the heavy-operator matrix elements are correlated, the Haar-averaged negativity may not be representative of the actual fixed PET state. The authors should either provide a randomness-independence argument or test the joint distribution numerically.","section":"Section III.A and III.B"},{"comment":"The main result is for an individual PET state, not for a Haar-averaged ensemble, so the variance suppression is load-bearing. The transition from Eq. (39) to the claim σ² ~ O(1/(D e^{S_2})) is only sketched: the counting factor S(m,n), the analytic continuation in m and n, and the final phase-space sum are not exhibited. A more detailed derivation or a quantitative bound on the higher moments is needed to justify replacing the Haar average ⟨N⟩_U by the actual N(ρ_R).","section":"Section III.A.2 (Variance)"}],"minor_comments":[{"comment":"The word 'resporting' should be 'resorting'.","section":"Section III.A, text before Eq. (27)"},{"comment":"The Wigner function is defined for qudits of odd prime dimension, but the SYK numerics in Appendix B use a qubit computational basis (d = 2, Hilbert space dimension 2^{N/2}). This does not invalidate the test of Assumption 1 as a statement about random eigenvectors, but it is not a direct test of the Wigner-negativity calculation; the authors should clarify whether the qubit test is intended only as heuristic support.","section":"Section II.C and Appendix B"},{"comment":"The Weingarten contraction formula contains index typos: the second line should involve δ_{j_a j'_{τ(a)}} for a = 1,...,2n, with the correct parentheses in the permutation labels.","section":"Eq. (29)"},{"comment":"The word 'eqaution' in the second paragraph should be 'equation'.","section":"Section IV, Discussion"},{"comment":"Reference [39] is cited only as 'Talk at Strings 2026'; this should be replaced by a published source or a more complete citation.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is built on a chain of explicit randomness assumptions and relies heavily on the authors' own previous work [20,30]; the incremental content beyond [20] is mostly the application to fixed-energy PET states and the geometric identification in Eq. (46). I would favor publication only if the load-bearing Assumption 1 is either substantially strengthened or the claims are appropriately weakened, since the current numerical test does not probe the full Haar structure used in the derivation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this paper computes a concrete magic monotone for the reduced density matrix of a fixed-energy PET state and gets a clean geometric answer, exp[(A_out - A_min)/(8G_N)]. The derivation is transparent and the formula is the kind of thing people in holographic complexity will want to check. Second thing: the whole calculation hangs on an assumption about the eigenbasis of a chaotic CFT being Haar-random with respect to the computational basis, and the numerical evidence offered only tests a piece of that assumption.\n\nWhat is actually new is the explicit application to PET states. The ingredients are known—Haar-random Wigner negativity, random tensor network Rényi entropies, and PET-state geometry—but putting them together for fixed-energy microcanonical windows is not in the earlier papers, and equation (46) is a real statement. The paper is upfront about what it is doing: it states Assumptions 1 and 2 explicitly, re-derives the negativity formula instead of just citing it, and includes an alternative integral-representation derivation in Appendix A that gives the same answer. The SYK numerics in Appendix B are a genuine attempt to test Assumption 1, and the distribution of <q|P|q> matches the Beta prediction for small windows. Credit where due: this is careful, readable work.\n\nThe soft spot is exactly where the stress-test note lands. Assumption 1 needs the full matrix U_{q,i} to be Haar-random: the negativity calculation uses Weingarten contractions that depend on the joint distribution of entries, including phases and higher-order correlations. Appendix B only checks the row-norm squares, x = <q|P|q>, through the Beta distribution. That statistic is insensitive to the very correlations that do the work in the derivation. So the SYK evidence is necessary but not sufficient. There is also a joint-statistics issue: in the actual CFT, rho_R and U are both functions of the same Hamiltonian, whereas the Haar average holds rho_R fixed. The paper does not address that coupling.\n\nThe variance suppression is argued heuristically; the scaling O(1/(D e^{S_2})) is plausible but not a proof. I would not call this a disqualifying issue—the paper is explicit about the conditional nature—but the main claim would be stronger if the randomness assumption were tested at the level of the full unitary, or if the authors narrowed the claim accordingly.\n\nWho should read it: anyone working on magic, stabilizer complexity, or the python's lunch in holography. It is a within-subfield note, not a breakthrough, but it is a solid calculation with a checkable output. It deserves peer review, with the referee focused on Assumption 1. I would not desk-reject it, and I would not cite it as established fact until the randomness assumption has better support. Send it out.","headline":"Clean formula for Wigner negativity in PET states, but the load-bearing Haar-randomness assumption is only tested at the level of marginals; worth review, not yet citable as established fact.","tokens_in":17592,"tokens_out":6308,"would_cite":false,"duration_ms":47245,"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 claims that the relative Wigner negativity of the right reduced density matrix of a fixed-energy PET state equals $\\exp[(A_{\\mathrm{out}}-A_{\\mathrm{min}})/8G_N]$, so a python's lunch exponentially enhances the boundary…","keywords":["stabilizer complexity","magic monotone","Wigner negativity","python's lunch","partially entangled thermal states","random matrix theory","holographic duality","quantum chaos"],"falsifier":"In the same random-coupling fermion model used in the paper's Appendix B, restrict to a microcanonical window narrower than the random-matrix scale and test the off-diagonal statistics of the overlap matrix $\\langle q|E_i\\rangle$, along with the predicted Wigner negativity $\\sqrt{2/\\pi}\\sqrt{D/e^{S_2(\\rho_R)}}$ of a fixed-energy PET state; a statistically significant deviation from the uniform-random-unitary prediction in either test would falsify the central formula.","tokens_in":16485,"feed_emoji":"🐍","tokens_out":14460,"duration_ms":120088,"temperature":0.7,"pith_summary":"The paper's aim is to show that a boundary notion of quantum complexity, the stabilizer complexity of a reduced density matrix, detects the python's lunch structure of a holographic wormhole geometry. In a partially entangled thermal (PET) state with fixed-energy boundary conditions, the paper computes the relative Wigner negativity of the right-side density matrix with respect to the microcanonical ensemble in the same energy window, and finds it to be $\\exp[(A_{\\mathrm{out}}-A_{\\mathrm{min}})/8G_N]$, where $A_{\\mathrm{out}}$ is the area of the outer extremal surface and $A_{\\mathrm{min}}$ the minimal one. Since Wigner negativity is a magic monotone, this quantity bounds the probability of preparing the state by stabilizer operations and therefore measures classical-simulation hardness. The result means the stabilizer complexity is $O(1)$ when the outer extremal surface is already minimal, but exponentially enhanced when a python's lunch is present.","feed_headline":"Python's lunch exponentially spikes stabilizer complexity","feed_subtitle":"A wormhole with a python's lunch makes the boundary state exponentially harder to prepare with stabilizer circuits.","key_machinery":"The machinery has three tiers. First, the discrete Wigner function $W_\\rho(u)=D^{-1}\\operatorname{Tr}(\\rho A(u))$ and its sum negativity define a magic monotone; the paper studies the ratio $\\mathcal{N}(\\rho|\\rho^{(0)})$ to remove the universal divergence with Hilbert-space dimension. Second, averaging over uniform random unitaries applied to the overlap matrix $U_{q,i}=\\langle q|E_i\\rangle$ turns the absolute-value problem into a replica computation in which only pairwise contractions survive at large $D$, yielding $N(\\rho_R)\\simeq\\sqrt{2/\\pi}\\sqrt{D/e^{S_2(\\rho_R)}}$. Third, the second-order entropy $S_2(\\rho)=-\\log\\operatorname{Tr}(\\rho^2)$ is evaluated by treating the state as a random tensor network, with $\\operatorname{Tr}\\rho_R^2=e^{-S(E_L^*)}+e^{-S(E_R^*)}$, so $S_2(\\rho_R)=\\min(S_L,S_R)$, which is the minimal area $A_{\\mathrm{min}}/4G_N$ at leading order. The ratio of negativities cancels the dimension $D$, leaving the area-gap exponential.","core_discovery":"On the paper's own terms, the discovery is a quantitative link between geometry and magic: for a fixed-energy PET state with pseudorandom Hamiltonian and Gaussian heavy-operator matrix elements, $N(\\rho_R|\\rho_R^{(0)}) \\sim \\exp[(A_{\\mathrm{out}}-A_{\\mathrm{min}})/8G_N]$. Here $\\rho_R$ is the right reduced density matrix, $\\rho_R^{(0)}$ is the maximally mixed microcanonical state in the same energy window, and the area gap is computed between the outer extremal surface of the right wedge and the smallest-area extremal surface. The relative Wigner negativity is basis-independent at the random-matrix energy scale, does not depend on the microscopic Hilbert-space dimension, and is exponentially large precisely when the entanglement wedge contains a python's lunch. Consequently the stabilizer complexity of the subregion is trivial without the lunch and exponentially enhanced with it.","pith_inferences":["If unitary-group randomness at the level of a two-design is sufficient, then the negativity formula should be reproducible with any operator basis that forms a two-design on the microcanonical subspace; this gives a concrete route to testing basis-independence in small chaotic systems.","The paper's numerical check only tests the diagonal distribution of $\\langle q|P|q\\rangle$; an off-diagonal test of the full overlap matrix $\\langle q|E_i\\rangle$ would probe the higher moments that enter the replica computation, and is likely where a breakdown of the randomness assumption would appear first.","If the fidelity bound is qualitatively tight, the formula converts the area gap into a statement about state preparation: any stabilizer protocol that prepares the right reduced state from the microcanonical ensemble must fail with probability close to one when $A_{\\mathrm{out}}-A_{\\mathrm{min}}$ is of order $G_N^{-1}$.","Projecting onto a microcanonical window at the random-matrix scale makes the relative time-shift mode highly uncertain, so it is worth checking whether the same area-gap formula survives when a semiclassical time-shift coherent state is used instead; the paper does not address this."],"forward_implications":["If the central formula holds, the stabilizer complexity of the right subregion is $O(1)$ whenever the outer extremal surface is already the minimal one, so a featureless black-hole wedge is computationally cheap to prepare from the microcanonical ensemble.","When a python's lunch is present, the relative Wigner negativity is $\\exp[(A_{\\mathrm{out}}-A_{\\mathrm{min}})/8G_N]$, and any stabilizer circuit that attempts to prepare $\\rho_R$ from the microcanonical state fails with probability at least $1-\\exp[-(A_{\\mathrm{out}}-A_{\\mathrm{min}})/8G_N]$.","The result is independent of the computational basis and of the microscopic Hilbert-space dimension, provided the energy window is at the random-matrix scale.","With appropriate pseudorandomness assumptions, the same argument extends to PET states with multiple operator insertions, fixed-area states, and multi-boundary black holes in three-dimensional gravity.","The quantity computed is a property of the boundary reduced state, not the bulk reconstruction map; its area dependence differs from the original python's lunch complexity formula, so the two should not be conflated."],"supporting_citations":[{"why":"Defines the python's lunch conjecture and the geometric complexity formula that this paper contrasts with its own result.","marker":"[7]"},{"why":"Supplies the leading large-dimension unitary-group averaging calculation for Wigner negativity of random states, including the contraction expansion and replica continuation used here.","marker":"[20]"},{"why":"Introduces partially entangled thermal states and the two-external-surface wormhole geometry that contains the python's lunch.","marker":"[21]"},{"why":"Provides a known formula for Wigner negativity of uniform random states that the paper imports.","marker":"[29]"},{"why":"Provides the bulk derivation of the matrix-element ansatz with pseudorandom coefficients used for the heavy insertion.","marker":"[32]"},{"why":"Establishes the random tensor network description of PET states and the second-order entropy formula used to compute $S_2(\\rho_R)$.","marker":"[34]"},{"why":"Gives the random-matrix energy scale in holographic theories that sets the microcanonical window for the main randomness assumption.","marker":"[38]"},{"why":"Formulates the eigenvector-randomness conjecture that the paper's first assumption extends to the many-body setting.","marker":"[60]"}],"fun_headline_variants":["Python's lunch in wormhole exponentially boosts magic","Stabilizer complexity goes exponential with a python's lunch","Python's lunch means exponential stabilizer cost for boundary state","Bulk python's lunch exponentially hardens boundary state preparation","Exponential magic from a bulk python's lunch"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire calculation rests on the premise that, inside a small energy window, the energy eigenstates are spread over the computational basis exactly as a random unitary drawn uniformly from the group of all unitaries would spread them; if that randomness fails, the negativity formula and its area interpretation do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Python's lunch in wormhole exponentially boosts magic","Stabilizer complexity goes exponential with a python's lunch","Python's lunch means exponential stabilizer cost for boundary state","Bulk python's lunch exponentially hardens boundary state preparation","Exponential magic from a bulk python's lunch"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001307,"raw_usage":{"total_tokens":5323,"prompt_tokens":933,"completion_tokens":4390,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":549,"completion_tokens_details":{"reasoning_tokens":4313}},"tokens_in":549,"tokens_out":4390,"duration_ms":26577,"temperature":1.0,"reasoning_tokens":4313,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:08:25.509231+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In the same random-coupling fermion model used in the paper's Appendix B, restrict to a microcanonical window narrower than the random-matrix scale and test the off-diagonal statistics of the overlap matrix $\\langle q|E_i\\rangle$, along with the predicted Wigner negativity $\\sqrt{2/\\pi}\\sqrt{D/e^{S_2(\\rho_R)}}$ of a fixed-energy PET state; a statistically significant deviation from the uniform-random-unitary prediction in either test would falsify the central formula.","supporting_citations":[{"cited_title":"Kitaev, A simple model of holography 2,https:// online.kitp.ucsb.edu/online/entangled15/kitaev2/ (2015)","cited_arxiv_id":null,"evidence_quote":"Formulates the eigenvector-randomness conjecture that the paper's first assumption extends to the many-body setting."}],"review_version":1}