{"id":"fcc50189-5a4c-4340-9aa6-0995888638a7","arxiv_id":"2411.12215","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Provides general recipes for constructing imaginarity measures via convex roofs and via the least pure-state imaginarity of input states, plus a no-go theorem for finite state-conversion criteria.","lead":"This paper builds two general ways to measure 'imaginarity', the part of a quantum state that cannot be described with real numbers alone. One method defines measures by mixing pure-state values; the other defines a measure as the least imaginarity of a pure state that can be converted into the target state by real operations.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The no-go theorem's functions f_k(x)=1-(x^k ∧ 1) are not concave for k∈(0,1), so the I_{f_k} measures are not established as valid imaginarity measures, leaving Theorem 2 unsupported.","rationale":"The reader's weakest_assumption correctly flags that Theorem 2's proof has an invalid inference from p2 a^k ≤ b^k to a ≤ b unless p2≥1, and that the parameter choices are not shown to be compatible. My independent reading finds a more basic defect in the same proof: the family f_k used to build the no-go monotones does not satisfy the concavity hypothesis of Theorem 1. Because the no-go theorem's monotonicity step relies entirely on I_{f_k} being a valid imaginarity measure, the theorem is unsupported even before the arithmetic issue. I do not see a similar defect in the main quantification results: Theorem 1's forward and converse arguments are consistent, Theorem 3's monotonicity proof follows the standard cost construction, and Theorem 5's qubit formula is derived correctly. Thus the paper's central quantification claim appears sound, but a headline application, the finite-measure no-go theorem, needs a corrected proof or a replacement family of genuinely concave functions. The reader's CONDITIONAL verdict remains appropriate, so I recommend no change to the verdict.","tokens_in":13933,"tokens_out":13894,"duration_ms":137979,"concrete_test":"Compute the concavity check for f_{1/2} at x=0.04, y=0.64, λ=1/2; if f(0.34)<0.5, the asserted concavity of f_k fails. Optionally, verify that I_{f_{1/2}} violates monotonicity under a real operation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 2's proof constructs measures I_{f_k} from f_k(x)=1-(x^k ∧ 1) and asserts that each f_k satisfies condition (iii), concavity. For x∈[0,1], x^k∧1 = x^k, so f_k(x)=1-x^k. For 0<k<1, x^k is concave, hence f_k is convex, not concave; the second derivative of f_k is k(k-1)x^{k-2}>0, confirming convexity. The displayed inequality in the text, (λa+(1−λ)b)∧1 ≥ λ(a∧1)+(1−λ)(b∧1), proves that g(x)=x∧1 is concave, which is the reverse of what is needed: substituting into f=1-g requires the inequality ≤ for g. A concrete violation: for k=1/2, x=0.04, y=0.64, λ=1/2, we get f(0.34)=1−√0.34≈0.417, while λf(x)+(1−λ)f(y)=0.5(1−0.2)+0.5(1−0.8)=0.5, so f(λx+(1−λ)y)<λf(x)+(1−λ)f(y). Thus f_k fails condition (iii). Consequently, Theorem 1 does not apply to f_k, and the claim that each I_{f_k} is an imaginarity measure is unproven. Since the monotonicity inequality in Theorem 2's proof is applied to these I_{f_k}, the no-go theorem lacks a valid proof. This is independent of the additional flaw that p2 a^k ≤ b^k does not imply a ≤ b when p2<1; both defects undermine the same theorem. The central quantification results (Theorems 1, 3, 4, 5) may still be correct, but the headline no-go application is not established as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops two approaches to quantifying imaginarity. First, for any decreasing concave f with f(1)=0, it defines I_f(rho) by the convex roof over pure-state values f(|<psi*|psi>|) and proves (Theorem 1) that I_f satisfies (I1)-(I5), with a converse stating that every imaginarity measure satisfying (I1),(I3),(I4) restricts to such an f on pure states. Second, for a state rho and the set R(rho) of pure states convertible to rho under real operations, it defines ~I_f(rho)=min_{|phi> in R(rho)} f(|<phi*|phi>|) and proves it is an imaginarity monotone (Theorem 3), gives a decomposition formula (Theorem 4), an analytic qubit expression (Theorem 5), and a characterization of convexity (Theorem 6). The paper also states a no-go theorem (Theorem 2) claiming that no finite set of imaginarity measures can decide mixed-state convertibility in dimension at least 4.","tokens_in":14386,"tokens_out":6016,"duration_ms":53637,"significance":"If the central theorems are correct, the paper provides a clean, unifying characterization of convex-roof imaginarity measures and a new operational monotone with a direct state-conversion interpretation, including an analytic qubit formula. The equivalence in Theorem 1 is elegant, and the proof strategy via Proposition 2 is largely self-contained. The numerical examples and the connection to geometric imaginarity add value. However, the no-go application in Theorem 2 is the advertised headline result and it is not currently established as written; the remaining theorems may still be sound, but the paper's main claim of a no-go theorem needs repair.","major_comments":[{"comment":"The functions f_k(x)=1-(x^k and 1) do not satisfy condition (iii) for 0<k<1. Since x^k is concave on [0,1] for k in (0,1), f_k is convex; the displayed inequality (lambda a+(1-lambda)b) and 1 >= lambda(a and 1)+(1-lambda)(b and 1) proves concavity of x and 1, which is the opposite of what is needed for f_k=1-(x and 1) to be concave. A concrete violation for k=1/2, x=0.04, y=0.64, lambda=1/2 is f_k(0.34) ≈ 0.417 < 0.5 = (f_k(x)+f_k(y))/2. Therefore Theorem 1 does not license the claim that each I_{f_k} is an imaginarity measure, and the monotonicity inequality used in the proof of Theorem 2 is unsupported.","section":"Section I, Theorem 2"},{"comment":"The step 'Choosing k = 1 - |<phi*_{eta0}|phi_{eta0}>|, one can get |<psi*_2|psi_2>| <= |<phi*_{eta0}|phi_{eta0}>|' does not follow. From the displayed inequality, using p1=1-p2, one obtains p2 |<psi*_2|psi_2>|^k <= |<phi*_{eta0}|phi_{eta0}>|^k; the inference to the unpowered inequality requires p2 >= 1, while p1>0 and p1+p2=1 imply p2<1. The proof also fails to show that the lower bound on p1 (needed for the existing measures I_j) can be satisfied simultaneously with the requirement that p2 be large enough for the contradiction; these two requirements may conflict. Theorem 2 is therefore not established as written.","section":"Section I, Theorem 2 proof"}],"minor_comments":[{"comment":"The word 'imagenarity' should be 'imaginarity'.","section":"Page 2"},{"comment":"The phrase 'for a tupe of quantum states' should be 'for a tuple of quantum states'.","section":"Page 5"},{"comment":"In statement (3), there is a mismatched parenthesis in 'f (sum_i ~p_i|<~psi*_i|~psi_i>|)) <= ...'; one parenthesis should be removed.","section":"Theorem 6"},{"comment":"The expression 'Phi(rho)=Phi(Phi_0(|psi>))' omits the state |psi><psi|; it should read Phi(rho)=Phi(Phi_0(|psi><psi|)).","section":"Proof of Theorem 3"},{"comment":"The definition of K_j contains a denominator sqrt(1-x), which is undefined when x=1; the case x=1 (all pure states real) should be handled separately or excluded explicitly, even though the statement can be recovered by a limiting argument.","section":"Proposition 2"}],"recommendation":"major_revision","confidential_remarks":"The paper's main structural results (Theorems 1, 3, 4, 5, and 6) appear sound and the exposition is readable. The problem is concentrated in Theorem 2, which is a headline no-go claim. If the authors can repair Theorem 2 by choosing valid concave functions and fixing the p2 inference, the paper would be a solid contribution. If not, the remaining results are still publishable but the paper's claim of a no-go theorem would need to be removed or substantially weakened. The reliance on the authors' own prior work [38, 41, 45] is appropriate and not problematic."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core quantification results are the real value here. Theorem 1 gives a clean characterization of convex roof imaginarity measures in terms of decreasing concave functions on [0,1], with a converse that is genuinely useful. Theorems 3–6 develop the least-imaginarity-of-input-pure-states monotone ~I_f, give it an operational reading as a formation cost under real operations, and provide a neat qubit formula. I checked the proofs of these theorems and they look sound, given the cited lemmas from the imaginarity and coherence literature. Table I is a handy summary. The paper is worth reading for these tools alone.\n\nThe soft spot is Theorem 2, the no-go result on mixed-state conversion. The proof constructs measures from f_k(x) = 1 - (x^k ∧ 1) for k ∈ (0,1]. For 0<k<1, x^k is concave, so f_k is convex, not concave. The inequality displayed to justify concavity, (λa+(1−λ)b)∧1 ≥ λ(a∧1)+(1−λ)(b∧1), shows the function x↦x∧1 is concave, but substituting f=1−g would require the reverse inequality. A simple counterexample with k=1/2 shows f_k violates concavity. Thus Theorem 1 does not apply to these f_k, and I_{f_k} is not established as an imaginarity measure. Independent of that, the proof of Theorem 2 also contains a faulty inference: from p2 a^k ≤ b^k with p2<1, the conclusion a ≤ b does not follow. Both defects strike the same theorem. The no-go claim might be true — a parallel result exists for general resource theories — but it is not proven here.\n\nThe paper would benefit from a careful revision of Theorem 2. The central characterization (Theorem 1) and the ~I_f monotone are separable and should survive intact. I would send this to a serious referee, because the main tools are valuable and mostly correct, and the flaws in Theorem 2 are isolated and fixable in principle, either by finding a genuinely concave family of functions for the no-go construction or by citing a different route. The authors should be asked to address the concavity issue and the p2 step directly.\n\nFor a reader working in resource theories of imaginarity or coherence, this is a useful addition. The flaws are real but contained. I'd bring it to a reading group to discuss the convex roof characterization, and I might cite it in my own work on imaginarity measures — though I would avoid relying on Theorem 2 until it is repaired.","headline":"The convex roof and least-input pure-state imaginarity constructions are solid and useful, but the no-go theorem (Theorem 2) is not proven as written because the f_k family is not concave and the final inequality step is invalid.","tokens_in":14919,"tokens_out":1566,"would_cite":true,"duration_ms":17837,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.65.Ud","03.67.-a","03.65.Ta"],"model":"deepseek-v4-flash","headline":"The paper proves that every imaginarity measure comes from a decreasing concave function of the pure-state overlap, and that no finite set of such measures decides mixed-state conversion under real operations.","keywords":["resource theory of imaginarity","convex roof construction","imaginarity measure","real operations","state conversion","decreasing concave function","qubit formula","imaginarity monotone"],"falsifier":"A concrete check of the no-go proof: the step from $p_2 a^k \\le b^k$ to $a\\le b$ is not valid for $0<p_2<1$; for example $a=0.9$, $b=0.5$, $k=0.5$, $p_2=0.5$ satisfies the premise while violating the conclusion, so the claimed contradiction is not forced.","tokens_in":13730,"feed_emoji":"🧮","tokens_out":13994,"duration_ms":130883,"temperature":0.7,"pith_summary":"The paper sets up a general framework for measuring imaginarity, the amount of a quantum state that genuinely requires complex numbers rather than real ones. It shows that any reasonable imaginarity measure on pure states is encoded by a single decreasing concave function $f$ of the overlap $|\\langle\\psi^*|\\psi\\rangle|$, and that extending $f$ to mixed states by the convex roof (the least average pure-state imaginarity over all decompositions) always yields a valid measure. It then introduces a second, operationally motivated quantifier: the least imaginarity of a pure state that a real operation can convert into the given mixed state. For qubits this quantifier has a closed form $f\\bigl(\\sqrt{1-4(\\operatorname{Im} b)^2}\\bigr)$ in terms of the off-diagonal element. Finally, the paper proves that no finite collection of imaginarity measures can completely determine which mixed states can be converted into which by real operations in dimension four or higher.","feed_headline":"All imaginarity measures come from one concave function","feed_subtitle":"Convex roofs turn any decreasing concave f into a valid measure, and qubit conversion costs are explicit.","key_machinery":"The central object is the conjugate overlap $x=|\\langle\\psi^*|\\psi\\rangle|$ of a pure state, together with a decreasing concave shape function $f(x)$ satisfying $f(1)=0$. Two constructions carry the argument: the convex roof $I_f(\\rho)=\\min_{\\sum_i p_i|\\psi_i\\rangle\\langle\\psi_i|=\\rho}\\sum_i p_i f(|\\langle\\psi_i^*|\\psi_i\\rangle|)$, and the conversion cost $\\tilde I_f(\\rho)=\\min_{|\\phi\\rangle\\in R(\\rho)} f(|\\langle\\phi^*|\\phi\\rangle|)$. The bridge between them is the fact that any pure-state ensemble can be produced from one canonical qubit pure state by a real operation, which turns the minimization over input states into a maximization over ensembles and yields the closed qubit formula.","core_discovery":"The central claim is that the whole family of convex-roof imaginarity measures is parameterized by decreasing concave functions $f:[0,1]\\to[0,1]$ with $f(1)=0$, with pure-state value $f(|\\langle\\psi^*|\\psi\\rangle|)$. The converse also holds: any imaginarity measure satisfying axioms (I1), (I3), and (I4) must restrict to pure states as such an $f$. The paper further constructs the monotone $\\tilde I_f(\\rho)=\\min_{|\\phi\\rangle\\in R(\\rho)} f(|\\langle\\phi^*|\\phi\\rangle|)$, where $R(\\rho)$ is the set of pure states convertible to $\\rho$ by real operations, and proves it is a valid imaginarity monotone equal to $f\\bigl(\\max_{\\{p_i,|\\phi_i\\rangle\\}}\\sum_i p_i |\\langle\\phi_i^*|\\phi_i\\rangle|\\bigr)$. For qubits the optimal decomposition is explicit, giving $\\tilde I_f(\\rho)=f\\bigl(\\sqrt{1-4(\\operatorname{Im} b)^2}\\bigr)$ with $b$ the off-diagonal element. Theorem 2 asserts that for dimension $d\\ge 4$ no finite set of imaginarity measures can classify mixed-state convertibility under real operations.","pith_inferences":["Taking Theorem 1's converse at face value, imaginarity is a one-dimensional resource on pure states, in contrast to coherence and entanglement where pure-state order is governed by majorization; this makes candidate measures comparable by a single scalar function.","The conversion-cost construction $\\tilde I_f$ looks like a one-shot formation cost for imaginarity; for well-chosen $f$ it may coincide with an asymptotic conversion rate, giving a second operational meaning beyond the geometric measure.","The qubit formula implies that for one qubit every monotone in this family is a monotone function of $|\\operatorname{Im} b|$, so they all impose the same state order; genuinely different orderings can only appear in dimension three or higher.","If the no-go theorem survives a repaired proof, it sharpens the general impossibility result for continuous faithful monotones by showing that even dropping those regularity requirements leaves an infinite hierarchy of conditions."],"forward_implications":["Any decreasing concave $f$ with $f(1)=0$ yields a valid imaginarity measure by convex roof, and every measure satisfying (I1), (I3), and (I4) is of this form on pure states.","The pure-state restriction of a convex-roof imaginarity measure depends only on $|\\langle\\psi^*|\\psi\\rangle|$, so pure states with equal conjugate overlap are equally imaginary under all such measures.","The quantifier $\\tilde I_f$ is a valid imaginarity monotone and is the largest monotone that agrees with $I_f$ on pure states; it equals $f\\bigl(\\max_{\\{p_i,|\\phi_i\\rangle\\}}\\sum_i p_i |\\langle\\phi_i^*|\\phi_i\\rangle|\\bigr)$.","For qubits, $\\tilde I_f(\\rho)=f\\bigl(\\sqrt{1-4(\\operatorname{Im} b)^2}\\bigr)$, where $b$ is the off-diagonal matrix element, making all these monotones explicitly computable on one qubit.","For dimensions $d\\ge 4$, no finite collection of imaginarity measures is complete for deciding mixed-state convertibility under real operations (Theorem 2)."],"supporting_citations":[{"why":"Defines the resource theory of imaginarity and the measure axioms (I1)-(I4) that the paper's constructions must satisfy.","marker":"[1]"},{"why":"Introduces real operations, geometric imaginarity, and the pure-state convertibility criterion that motivates the conversion-cost construction.","marker":"[15]"},{"why":"Supplies the canonical form of pure states under real unitaries (Proposition 1) and the qubit mixed-state convertibility criteria that Theorem 2 extends.","marker":"[28]"},{"why":"Introduces condition (I5) and the equivalence between (I3)+(I4) and (I2)+(I5), used in the converse part of Theorem 1.","marker":"[29]"},{"why":"Establishes that every state can be produced from some pure input by real operations, which defines the set R(ρ) underlying tilde I_f.","marker":"[32]"},{"why":"Shows no finite complete set of continuous faithful resource monotones exists in general resource theories, the background result that Theorem 2 sharpens for imaginarity.","marker":"[42]"},{"why":"Provides the parallel coherence result that the mixed-state extension via least pure-state coherence is the supremum of monotones agreeing on pure states.","marker":"[49]"},{"why":"Supplies the convex-roof method for coherence that is adapted here to imaginarity.","marker":"[38]"}],"fun_headline_variants":["Imaginarity measures: one concave function to rule them all","Convex roofs and pure states fully characterize imaginarity","Qubit imaginarity: explicit formula from optimal decomposition","No finite set of imaginarity measures can classify conversions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The no-go theorem's construction assumes that the mixing probabilities in the prepared state can be chosen so that every finite set of measures sees the mixture as at least as imaginary as the target while the specially constructed measure sees it as strictly less imaginary; the proof never demonstrates those two requirements are compatible.","fun_headline_variants_meta":{"raw":{"variants":["Imaginarity measures: one concave function to rule them all","Convex roofs and pure states fully characterize imaginarity","Qubit imaginarity: explicit formula from optimal decomposition","No finite set of imaginarity measures can classify conversions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000586,"raw_usage":{"total_tokens":2738,"prompt_tokens":915,"completion_tokens":1823,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":531,"completion_tokens_details":{"reasoning_tokens":1753}},"tokens_in":531,"tokens_out":1823,"duration_ms":14469,"temperature":1.0,"reasoning_tokens":1753,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T17:49:38.885363+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check of the no-go proof: the step from $p_2 a^k \\le b^k$ to $a\\le b$ is not valid for $0<p_2<1$; for example $a=0.9$, $b=0.5$, $k=0.5$, $p_2=0.5$ satisfies the premise while violating the conclusion, so the claimed contradiction is not forced.","supporting_citations":[{"cited_title":"In addition, if f is strictly decreasing, then If is faithful, that is, If (ρ) = 0 if and only if ρ is a real state","cited_arxiv_id":null,"evidence_quote":"Defines the resource theory of imaginarity and the measure axioms (I1)-(I4) that the paper's constructions must satisfy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces real operations, geometric imaginarity, and the pure-state convertibility criterion that motivates the conversion-cost construction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces condition (I5) and the equivalence between (I3)+(I4) and (I2)+(I5), used in the converse part of Theorem 1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes that every state can be produced from some pure input by real operations, which defines the set R(ρ) underlying tilde I_f."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows no finite complete set of continuous faithful resource monotones exists in general resource theories, the background result that Theorem 2 sharpens for imaginarity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the parallel coherence result that the mixed-state extension via least pure-state coherence is the supremum of monotones agreeing on pure states."}],"review_version":1}