{"id":"58b203ca-c928-4d80-8eae-c575f009faf3","arxiv_id":"1908.08514","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"A claimed identification of Virasoro circuit complexity with the Berry connection fails a basic consistency check for pure rotations.","lead":"This paper claims that for 2D conformal field theories, the cost of a quantum circuit built from symmetry gates equals the Berry connection, leading to a log-overlap measure of complexity. The central comparison of two known formulas contains a sign inconsistency that breaks the derivation.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sign check on the rotation protocol falsifies Eq. (47): the Virasoro Berry connection and the complexity have the same sign, not opposite.","rationale":"I reproduced the reader's sign check by direct substitution into the paper's own equations. For a rotation path with the vacuum state, Eq. (45) and Eq. (24) give identical positive expressions, ∫_g A_{h,c}=C_{h,c}=cθ/24, so the claimed relation (47), ∫_g A_{h,c}=−C_{h,c}, is false as an algebraic identity inside the paper's framework. This is not a subtle issue of conventions: the signs of both displayed formulas are fixed by the paper, and the rotation is a permitted protocol satisfying the stated path/projection conditions. In fact, the correct sign ∫_g A=+C is also what is needed for the generalized Berry phase (43) to vanish for pure rotations, as it should for a path whose projection to the quotient is trivial. The paper's Eq. (50) adds a further questionable step by replacing the signed endpoint phase with −|log⟨ψ_R|ψ_T⟩|, but the primary defect is the sign in Eq. (47). Because the central result fails on the simplest explicit example, the derivation of the complexity–Berry-phase proportionality and the logarithmic overlap formula is unsupported. The reader's REJECT verdict is therefore appropriate; my stress-test does not change it.","tokens_in":20892,"tokens_out":7887,"duration_ms":73427,"concrete_test":"Evaluate the claimed equality (47) for the explicit rotation protocol g(s,σ)=σ+(s/τ)θ with reference state |h=0⟩. From Eq. (45) compute ∫_g A_{0,c}=cθ/24; Eq. (46) is identically zero for rotations. From Eq. (24) compute C_{0,c}[g]=cθ/24. The equality (47) would require cθ/24=−cθ/24, which fails for any nonzero θ. As a cross-check, compute the generalized Berry phase (43) for the same path: B=cθ/24−i log(e^{-icθ/24})=0, consistent with a pure rotation being an element of the stabilizer; the paper's sign in Eqs. (47)–(48) instead gives −cθ/12.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central identification (47) is already false for the simplest admissible protocol, a rigid rotation g(s,σ)=σ+(s/τ)θ with the vacuum reference state |h⟩=0. Inserting this path into the Berry pieces: Eq. (45) gives ∫_g A_{h,c}=+cθ/24 (with ∫dσ=2π), while Eq. (46) vanishes because g''=0. The Virasoro complexity in Eq. (24) gives C_{h,c}[g]=+cθ/24, since Sch[g,σ]=0 and 12h/c=0. Thus the paper's own formulas give ∫_g A_{h,c}=+C_{h,c}[g], not −C_{h,c}[g]. The large-c limit cannot repair this: both sides are linear in c, and their relative sign does not depend on c. The same rotation protocol also exposes a consistency check via Eq. (43): the generalized Berry phase is B=∫_g A_{h,c}−i log⟨h|U[g^{-1}(0)g(τ)]|h⟩=cθ/24−i log(e^{-icθ/24})=0, which is the expected vanishing holonomy for a path lying in the stabilizer. The sign choice in Eq. (47), and hence in Eq. (48), would instead give a nonzero value. The subsequent derivations of Eqs. (50), (52), and (53) therefore rest on an algebraic sign error in comparing two formulas already present in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper investigates a relation between circuit complexity defined on Virasoro group manifolds and Berry phases in unitary representations of the Virasoro group. After reviewing the geometric complexity framework of Caputa and Magan [61] and the group-theoretic Berry phase construction of Oblak [75], the paper proposes that in the large central charge limit the Virasoro computational cost coincides with the negative of the Virasoro Berry connection (Eq. (47)). From this it derives that the generalized Berry phase equals minus the complexity plus an endpoint term (Eq. (48)), and then, for rotation-type protocols, a proportionality between full CFT complexity and -log|overlap|^2 (Eqs. (50), (52), (53)). The final discussion suggests a connection between the Berry phase and the Liouville action in the path-integral optimization proposal.","tokens_in":21263,"tokens_out":9489,"duration_ms":86481,"significance":"If valid, the identification would link three independent notions—circuit complexity, Berry phase, and state overlap—in two-dimensional CFTs, and would provide a symmetry-circuit origin for logarithmic complexity formulas that mimic holographic and path-integral complexity predictions. The paper has positive features: it uses no free parameters, it builds on explicit formulas from Ref. [61] and Ref. [75], and the final proportionalities are concrete and falsifiable. The central identification, however, fails a direct check using the paper's own equations, and the subsequent derivations do not survive that failure.","major_comments":[{"comment":"The claimed equality ∫_g A_{h,c} = -C_{h,c}[g](τ) is already false for the simplest admissible protocol. Take the rigid rotation path g(s,σ) = σ + (s/τ)θ with reference state |h>. Then g' = 1, g'' = 0, and Sch[g,σ] = 0. Eq. (45) gives ∫_g A_{h,c} = θ(c/24 - h) (the second piece, Eq. (46), vanishes), while Eq. (24) gives C_{h,c}[g] = θ(c/24 - h). The paper's own formulas therefore give ∫_g A_{h,c} = +C_{h,c}[g], not -C_{h,c}[g]. The relative sign is independent of c, so the large-c limit cannot repair the identification. Since Eqs. (48), (50), (52), and (53) all inherit this sign, the central claim is unsupported.","section":"V.B, Eq. (47)"},{"comment":"The sign error is not merely aesthetic; it contradicts the paper's own vanishing-holonomy statement. For the same rotation path, Eq. (49) gives U_θ|h> = exp(iθ(h - c/24))|h>, so Eq. (43) yields ∫_g A_{h,c} - i log<h|U[...]|h> = θ(c/24 - h) + θ(h - c/24) = 0. With Eq. (47) instead the right-hand side becomes -2θ(c/24 - h) ≠ 0 (for h ≠ c/24), in conflict with the statement in Sec. IV.B that the generalized Berry phase vanishes for any path of the form g(s) = g(0)h(s) with h(s) ∈ G_φ. The rotation protocol is exactly such a path.","section":"V.A/V.B, Eqs. (43) and (50)"},{"comment":"The replacement of -i log<ψ_R|ψ_T> by -|log<ψ_R|ψ_T>| is not an algebraic identity and is false for the rotation protocol. For the state in Eq. (49), |<ψ_R|ψ_T>| = 1, so -|log<ψ_R|ψ_T>| = 0 if the logarithm of the modulus is meant, whereas the boundary term in Eq. (48) equals θ(h - c/24); if the absolute value of the complex logarithm is meant, the sign and magnitude generally do not match. Thus Eqs. (52) and (53) are not derived consequences of Eq. (48).","section":"V.C, Eq. (50)"}],"minor_comments":[{"comment":"The Schwarzian is written as g'''/g' - (√3 g'')^2/(√2 g')^2; this is correct but needlessly opaque, and writing (3/2)(g''/g')^2 would improve readability.","section":"III.A, Eq. (17)"},{"comment":"The text contains the typo 'Mauer-Cartan'; it should read 'Maurer-Cartan' in both occurrences.","section":"IV.B"},{"comment":"In the Introduction, 'under a diabatic variations' should presumably be 'under adiabatic variations'; the terminology is otherwise standard throughout the paper.","section":"I, near Eq. (27)"}],"recommendation":"reject","confidential_remarks":"This is a clear-cut rejection: the sign error in Eq. (47) and the invalid replacement leading to Eq. (50) undermine all main results. A local sign flip does not fix the paper, because the Berry-phase holonomy would then have the wrong sign; the comparison of formulas must be reworked, and the main conclusions may change. I see no concern about novelty or citation ethics."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe main result of this paper is wrong. The central identification, Eq. (47), fails a two-line check: for a pure rotation of the vacuum, the Berry connection equals the Virasoro complexity, not its negative. Everything that follows—Eqs. (48), (50), (52), (53)—inherits that sign error.\n\nI'll give credit where it's due. The paper is readable and does a serviceable job of reviewing the Caputa-Magan cost function and Oblak's Berry phase in Virasoro representations. The discussion in Sec. VI about why the log-overlap formula is not a sensible complexity measure in discrete systems is honest and useful. The citations to the holographic complexity and path-integral optimization literature are appropriately placed.\n\nThe soft spot is the sign. Take g(s,σ)=σ+(s/τ)θ and the vacuum h=0. Eq. (45) gives +cθ/24, Eq. (46) vanishes, and Eq. (24) gives C=+cθ/24. So ∫ A = +C for the simplest admissible protocol. The large-c limit does not help because both sides scale as c and the sign is independent of c. The consistency check in Eq. (43) confirms this: the generalized Berry phase vanishes for a path in the stabilizer, but the paper's Eq. (48) with the wrong sign would give a nonzero value. This is not a convention or a typo; it is an algebraic mismatch between two formulas the paper itself reproduces.\n\nThe later proportionality C ∝ -log|⟨ψ_R|ψ_T⟩|² is therefore unsupported. The paper's own caveats in Sec. VI about the log formula are appropriate, but they don't fix the broken step. The citation pattern is fine, and the writing is clear—but this is a reflection with a load-bearing error, not a result.\n\nRecommendation: reject, and I wouldn't send it to a referee. A reader interested in the ingredients can go to the primary sources. A reading group could use it as a case study in sign bookkeeping, but there are better uses of time.","headline":"The central equality (47) has the wrong sign, and the paper's main claims fall with it.","tokens_in":21709,"tokens_out":5195,"would_cite":false,"duration_ms":47482,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In the large central charge limit, the paper claims that Virasoro circuit complexity is minus the integrated Berry connection, and derives from this a log-overlap formula for complexity in two-dimensional CFTs.","keywords":["circuit complexity","Berry phase","Virasoro group","two-dimensional conformal field theory","large central charge","coadjoint orbits","geometric action","Liouville action"],"falsifier":"Evaluate both sides of Eq. (47) for the rotation protocol $g(s,\\sigma)=\\sigma+s\\theta$ with the SL(2,$\\mathbb{R}$)-invariant vacuum as reference state: Eq. (45) contributes $+c\\theta/24$, while $-C_{0,c}$ from Eq. (24) contributes $-c\\theta/24$. This direct mismatch for the simplest allowed protocol is a concrete falsifier of the claimed identity.","tokens_in":20691,"feed_emoji":"🌀","tokens_out":14384,"duration_ms":128280,"temperature":0.7,"pith_summary":"This paper aims to establish that, in two-dimensional conformal field theories, the computational cost of a circuit built from Virasoro symmetry gates is the negative of the integrated Berry connection of the Virasoro group's unitary representation, at least in the large central charge limit. From that identification it derives a relation between the Virasoro Berry phase, the circuit complexity, and the state overlap: the Berry phase equals minus the complexity minus $|\\log\\langle\\psi_R|\\psi_T\\rangle|$, so that the complexity is proportional to $-\\log|\\langle\\psi_R|\\psi_T\\rangle|^2$. A sympathetic reader would care because this turns a logarithmic, state-dependent complexity formula that had previously been proposed on heuristic grounds into a corollary of a concrete circuit construction, and it connects complexity with geometric actions and the Liouville action that appear in holography.","feed_headline":"Virasoro circuit complexity is minus the Berry connection integral","feed_subtitle":"In the large central charge limit, the Berry phase and state overlap determine the cost of a Virasoro symmetry circuit.","key_machinery":"The carrying object is the Virasoro group manifold, i.e. the centrally extended group of orientation-preserving circle diffeomorphisms, used as the space of circuits: a protocol is a path $g(s,\\sigma)$ in the group, and the complexity is the integrated cost functional (24), $C_{h,c}[g](\\tau) = \\frac{c}{24\\pi}\\int_0^\\tau ds \\int_0^{2\\pi} d\\sigma\\, \\frac{\\dot{g}}{g'}\\left(\\frac12 - \\frac{12h}{c} + \\mathrm{Sch}[g,\\sigma]\\right)$, which coincides with the Kirillov geometric action on coadjoint orbits. The Berry connection is $A_{h,c} = i\\langle h | u[\\hat{\\Theta}_g] | h\\rangle$, the expectation value of the centrally extended Maurer-Cartan form in a highest-weight representation built on a primary state of weight $h$ and central charge $c$; it splits into a centreless term (45) and a Bott-cocycle term, i.e. the two-cocycle that centrally extends the circle-diffeomorphism group (46), and the load-bearing step is the claim that their sum equals the negative of the cost integrand. The generalized Berry phase for open paths adds a boundary contribution from the stabilizer of the highest-weight state, which is what converts the endpoint term into $|\\log\\langle\\psi_R|\\psi_T\\rangle|$.","core_discovery":"The paper's central claim is Eq. (47): in the large central charge limit, $\\int_g A_{h,c} = - C_{h,c}[g](\\tau)$, where the left side is the integral of the Virasoro Berry connection along the protocol path in the group manifold and the right side is the Virasoro circuit complexity for the same path. Using the generalized Berry phase for open paths whose endpoints differ by a rotation, the author rewrites the full Berry phase as $B_{h,c}[g](\\tau) = -C_{h,c}[g](\\tau) - |\\log\\langle\\psi_R|\\psi_T\\rangle|$, Eq. (50). Combining left and right chiral sectors gives the proportionality $C_{\\mathrm{CFT}} \\propto -\\log|\\langle\\psi_R|\\psi_T\\rangle|^2$, Eq. (53). The paper frames this as showing that the Berry phase of a primary state transported by Virasoro transformations is, up to an endpoint term fixed by the reference-target overlap, just the negative of the symmetry-circuit complexity.","pith_inferences":["Beyond the paper, if Eq. (50) survives, a state-overlap complexity measure must secretly know the gate set: it should be trusted only when the underlying circuit is made of Virasoro symmetry gates, not as a universal state distance.","Beyond the paper, the identification of complexity with $-\\int A_{h,c}$ suggests complexity inherits gauge freedom from the Berry connection; rephasing the reference or target states should shift the endpoint term, a concrete prediction that can be checked within the same group-manifold formalism.","Beyond the paper, a natural testable extension is to take non-rotation paths, such as special conformal transformations, and check whether the endpoint term in Eq. (50) remains path-independent; if it varies, the formula holds only for rotation protocols.","Beyond the paper, one could compare the time evolution of $-\\log|\\langle\\psi_R|\\psi_T\\rangle|^2$ after a global quench with known holographic complexity growth; agreement would suggest that holographic complexity probes Virasoro Berry phases."],"forward_implications":["In large-$c$ 2D CFTs, circuit complexity built from Virasoro symmetry gates becomes computable as a group-theoretic holonomy, since it is (minus) the integrated Berry connection along the protocol path.","The Virasoro Berry phase is fixed by Eq. (50) to be the negative complexity plus a path-independent endpoint term, so measuring the Berry phase is equivalent to measuring the complexity up to a constant.","The logarithmic complexity formula $C_{\\mathrm{CFT}}\\propto-\\log|\\langle\\psi_R|\\psi_T\\rangle|^2$ follows from a concrete circuit construction with symmetry gates, rather than being a standalone state-distance proposal.","The same relations suggest a proportionality between the Virasoro Berry phase and the classical Liouville action, connecting the Berry-phase formalism to the path-integral optimization measure of complexity.","Because the derivation is for continuous CFT systems, the paper predicts the logarithmic formula is not a valid complexity measure for discrete qubit systems; applying it there yields infinite complexity for a single-qubit flip."],"supporting_citations":[{"why":"Introduces circuit complexity defined on symmetry group manifolds and supplies the group-theoretic protocol framing used throughout.","marker":"[51]"},{"why":"Provides the Virasoro cost function and complexity functional in Eq. (24) that the paper equates with the negative Berry connection.","marker":"[61]"},{"why":"Supplies the Virasoro Berry connection formulas in Eqs. (45)-(46) and the group-theoretic Berry phase construction used for Eq. (43).","marker":"[75]"}],"fun_headline_variants":["Virasoro complexity equals minus Berry connection","Berry phase computes Virasoro circuit complexity","Complexity from Virasoro Berry holonomy","Virasoro circuits: cost from Berry connection","Negative Berry connection gives Virasoro complexity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation stands on the sign-matching of Eqs. (45) plus (46) to the negative of the expression for the circuit cost, and this matching already fails for the simplest allowed protocol: rotating the vacuum state gives $+c\\theta/24$ on the Berry side and $-c\\theta/24$ on the complexity side.","fun_headline_variants_meta":{"raw":{"variants":["Virasoro complexity equals minus Berry connection","Berry phase computes Virasoro circuit complexity","Complexity from Virasoro Berry holonomy","Virasoro circuits: cost from Berry connection","Negative Berry connection gives Virasoro complexity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000308,"raw_usage":{"total_tokens":1817,"prompt_tokens":1059,"completion_tokens":758,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":675,"completion_tokens_details":{"reasoning_tokens":690}},"tokens_in":675,"tokens_out":758,"duration_ms":7754,"temperature":1.0,"reasoning_tokens":690,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:40:23.029826+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate both sides of Eq. (47) for the rotation protocol $g(s,\\sigma)=\\sigma+s\\theta$ with the SL(2,$\\mathbb{R}$)-invariant vacuum as reference state: Eq. (45) contributes $+c\\theta/24$, while $-C_{0,c}$ from Eq. (24) contributes $-c\\theta/24$. This direct mismatch for the simplest allowed protocol is a concrete falsifier of the claimed identity.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Virasoro Berry connection formulas in Eqs. (45)-(46) and the group-theoretic Berry phase construction used for Eq. (43)."}],"review_version":1}