{"id":"d4c152f7-1be6-450c-bca6-16293dc7eb99","arxiv_id":"2607.19070","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Consumable magic states can simulate non-classically-simulable measurements and improve state discrimination, while finite-dimensional quantum catalysts and memories cannot improve discrimination of positive-Wigner state pairs.","lead":"This paper shows that consumable magic states can strengthen measurements that are classically simulable, but reusable quantum catalysts and memories cannot. It gives a sharp qutrit example where simulation cost is exactly one, and a no-go theorem for positive-Wigner state pairs.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3 rests on an undefined branch-wise CPWP-instrument property; without it, the conditional memory state may not be PWF and the supermartingale bound collapses.","rationale":"I agree with the reader's identified weakest assumption: the proof of Theorem 3 depends on the conditional memory state being PWF after every history, and this is asserted without a formal definition of 'CPWP instrument' or a proof of branch-wise PWF preservation. The concern is not merely cosmetic — the per-round bound and the supermartingale argument fail if conditioning can produce Wigner negativity. This is the single most load-bearing issue. The rest of the paper appears internally sound: Theorems 1 and 2 and the qutrit example are arithmetically consistent, and the reduction to distinguishing µ_C from γ_C is otherwise valid. A secondary overstatement is that the abstract and conclusion say 'success probability' where the theorem actually establishes equality of asymptotic success rates, but this does not undermine the mathematical core. Since the reader already assigned a CONDITIONAL verdict with this exact concern, my stress-test does not move the verdict; it sharpens the required fix: formalize branch-wise CPWP instruments and prove the conditional-state preservation step, or add a counterexample check.","tokens_in":20576,"tokens_out":9947,"duration_ms":110863,"concrete_test":"Perform a computational search over qutrit instruments: fix a PWF state ρ⊗γ (for example, computational basis states) and enumerate two-outcome instruments {Λ_0, Λ_1} whose total map Λ_0+Λ_1 is CPWP, but test whether an individual branch Λ_y can map a PWF input to an unnormalized state with negative discrete Wigner function. Compute W(Tr_A[Λ_y(ρ⊗γ)]) using the phase-space-point-operator expansion. If any such branch exists, the assertion that γ_C(F_j) is always PWF is false and Theorem 3's proof needs an added branch-wise CPWP assumption. If no such counterexample exists, state and prove the branch-wise preservation lemma; that would patch the proof.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The no-go theorem (Theorem 3, Section IV) reduces R_m ≤ p* to Eq. (103): for every history F_j, Pr(X_{j+1}=1|F_j) ≤ p*. This reduction uses Lemma 1, which requires the conditional memory state γ_C(F_j) to be PWF. The proof asserts this in one sentence: 'Since ρ0, ρ1, γ_C are PWF states, and each instrument is CPWP instrument, the conditional memory state γ_C(F_j) is a PWF state.' But the paper defines CPWP channels (Eq. 9) and CPWP-preserving superchannels (Eq. 10), not CPWP instruments. A quantum instrument is a collection of CP trace-non-increasing maps Λ_y whose sum is a channel. Knowing only that the sum map is CPWP does not imply each branch Λ_y preserves Wigner positivity; indeed, a sum of non-positive maps can be positive. After conditioning on outcome y, the updated memory state is (up to normalization) Tr_A[Λ_y(ρ⊗γ_C(F_j))]. If Λ_y creates Wigner negativity from a PWF input, γ_C(F_j) need not be PWF, Lemma 1 cannot be applied, and the supermartingale argument fails. The proof needs either a formal definition of 'CPWP instrument' that includes branch-wise PWF preservation, or a proof that such preservation follows from the definition. This is directly load-bearing because Eq. (102) and Eq. (103) are the only place where the PWF assumption on the inputs is used to control the γ_C-hypothesis in the charging argument.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies binary quantum state discrimination under positive-Wigner-function (PWF) measurements, i.e., classically simulable measurements (CSMs), in odd-prime-dimensional systems. It proposes three potential ways to improve the discrimination power of CSMs: adding consumable magic states, using quantum catalysts, and using quantum memories. The main technical contributions are: (i) a correspondence between PWF POVMs and completely positive Wigner-preserving (CPWP) measurement channels (Theorem 1); (ii) lower and upper bounds on the exact simulation cost of a non-CPWP measurement channel with a pure magic state (Theorem 2); (iii) a tight example using the qutrit Strange state for which the simulation cost is exactly one (Proposition 1); (iv) an example showing that one copy of the Strange state improves the optimal success probability over CSMs (Proposition 2); (v) an SDP formulation for magic-assisted discrimination; and (vi) a no-go theorem claiming that finite-dimensional quantum memories and catalysts do not improve the optimal success probability for discriminating a pair of PWF states (Theorem 3 and Corollary 1). The paper concludes that consumable magic can help while reusable magic in the form of memories or catalysts cannot, at least for PWF state pairs.","tokens_in":20984,"tokens_out":12425,"duration_ms":131552,"significance":"If the no-go theorem is correct, the paper establishes a clean conceptual separation between consumable and reusable magic resources in the context of quantum state discrimination, and it provides a useful channel-simulation perspective on CSMs. Theorems 1 and 2 and Propositions 1 and 2 appear arithmetically sound and are presented constructively, with explicit Wigner-function calculations; the tight Strange-state example is a valuable demonstration that the proposed bounds can coincide. The SDP formulation and its dual are also useful tools. The central weakness is the proof of Theorem 3, which depends on an unstated and unproved assumption about 'CPWP instruments.' Because the no-go result is the paper's headline claim, its soundness must be established before the result can be accepted.","major_comments":[{"comment":"The proof hinges on the assertion that the conditional memory state γ^C(F_j) is PWF 'since ρ0, ρ1, γ_C are PWF states, and each instrument is CPWP instrument.' However, the paper defines CPWP channels (Eq. (9)) and CPWP-preserving superchannels (Eq. (10)), but never defines a 'CPWP instrument.' For a quantum instrument, knowing that the sum map is CPWP does not imply that each branch Λ_y preserves Wigner positivity: a sum of maps that individually create Wigner negativity can itself be PWF-preserving. After conditioning on F_j and on the current outcome, the unnormalized updated memory state is proportional to Tr_A[Λ_y(ρ_i⊗γ^C(F_j))]; if Λ_y does not preserve PWFs, γ^C(F_j) need not be PWF. Lemma 1 then cannot be applied in Eq. (102), and the per-round bound Pr(X_{j+1}=1|F_j) ≤ p^* in Eq. (103) does not follow. The proof needs either a formal definition of 'CPWP instrument' that includes","section":"Section IV, proof of Theorem 3, Eqs. (102)-(103)"},{"comment":"Because the proof of Theorem 3 relies on the unproved branch-wise CPWP-instrument assumption, the unconditional statements in the abstract and conclusions—'neither finite-dimensional quantum memories nor quantum catalysts can improve the optimal success probability'—are stronger than what is proved. At minimum, Theorem 3 should be stated as a theorem about protocols whose adaptive instruments preserve Wigner positivity branchwise, with the assumption explicit. If the authors instead intend 'CPWP instrument' to mean exactly that branch-wise property, the term must be defined and the framework in Fig. 2/Definition 6 must be built on that definition. As written, the no-go claim is not established for arbitrary finite-dimensional memory-assisted protocols.","section":"Abstract, Section IV, Section V"}],"minor_comments":[{"comment":"There is a formatting typo: 'Tu Bτ−klZkXl' should presumably be 'T_u = τ^{-kl} Z^k X^l'.","section":"Eq. (3)"},{"comment":"The phrase 'odd-prime-dimensional system A with odd-prime number d_B' is confusing; it should read that the output system B has odd-prime dimension d_B.","section":"Theorem 1 statement"},{"comment":"The notation C^ε_F in Definition 5 is used with an approximation parameter ε, but Theorem 2 uses C_A(M;ω) without ε. Please define the exact simulation cost as a separate object and keep the notation consistent.","section":"Definition 5 and Theorem 2"},{"comment":"The term 'adaptive CPWP instrument' is used operationally before being formally defined. Even aside from the branch-wise issue raised above, a precise definition of the instrument, its outcomes, and the update rule for the memory is needed.","section":"Section IV, before Definition 6"},{"comment":"The sentence 'this example shows that the assistance of a single copy of the Strange state is still insufficient to perfectly distinguish...' is not established by the displayed calculation: the proof gives a lower bound 7/9, not an upper bound below 1. Either cite an external result for the insufficiency or rephrase the sentence as a weaker observation/conjecture.","section":"Proposition 2, paragraph after Eq. (76)"},{"comment":"The statement that the preparation channel P^{A→AC}_{γ^C} is CPWP 'for any PWF input state ρ^A' is slightly imprecise: CPWP is defined with an arbitrary reference system, so the justification should explicitly say that W_{ρ⊗γ}(u,v)=W_ρ(u)W_γ(v) for any joint PWF input.","section":"Lemma 1 proof"}],"recommendation":"major_revision","confidential_remarks":"The core technical issue is the undefined 'CPWP instrument' assumption in Theorem 3. I do not think this is a case for rejection, because the gap is localized and may be fixable by adding a formal definition and either proving the branch-wise preservation property or restating the theorem for that restricted class. However, the no-go theorem is the main advertised contribution, so the revision must address this point directly. The rest of the paper—Theorems 1-2 and Propositions 1-2, including the tight Strange-state example and the SDP—appears sound and publishable if the no-go part is repaired or appropriately qualified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the channel-simulation half of this paper is good and the arithmetic checks out. The memory/catalyst no-go theorem, which is the headline result, has a real gap in the proof, and the abstract overstates what is proven.\n\nWhat's actually new: Theorem 1 (PWF POVM iff its measurement channel is CPWP) is a clean equivalence, proved properly via the Choi matrix. Theorem 2's simulation bounds in terms of mana are a natural extension of the authors' own earlier work, and the qutrit Strange-state example (Props 1 and 2) is carefully constructed: I checked the Wigner-function values and the bounds coincide correctly, and the demonstration that one copy of the Strange state strictly raises the CSM success probability from 2/3 to 7/9 is a nice, explicit illustration. The SDP dual in the appendix is standard but usable.\n\nThe soft spot is Theorem 3. The proof needs the conditional memory state γ_C(F_j) to be PWF for every history, and the only justification is the sentence 'since ρ0, ρ1, γ_C are PWF states, and each instrument is CPWP instrument, the conditional memory state γ_C(F_j) is a PWF state.' But the paper defines CPWP channels and superchannels, not CPWP instruments. An instrument is a set of CP maps whose sum is a channel; knowing the sum is CPWP does not imply each branch preserves Wigner positivity. A sum of non-positive maps can easily be positive. After conditioning on an outcome, the memory update uses one branch, so γ_C(F_j) can leave the PWF set. If that happens, Lemma 1 doesn't apply, Eq. (102) and (103) don't follow, and the supermartingale argument collapses. This is the load-bearing step of the no-go claim, not a technical aside. The fix might be straightforward — define 'CPWP instrument' to mean each branch is CPWP, or prove the needed property some other way — but as written the theorem is unproven.\n\nThere is also a smaller mismatch: Theorem 3 is about the asymptotic success rate R_m, but the abstract and conclusion say 'optimal success probability'. The theorem's own statement is precise; the front matter overstates it.\n\nBottom line: the paper deserves peer review, because the channel-simulation part is solid and the no-go theorem is plausible and potentially important. But a referee needs to push hard on the instrument definition and the conditional-state argument, and the authors should align the abstract with what the proof actually establishes. I'd be interested in seeing the revised version.","headline":"Solid channel-simulation results with a tight Strange-state example, but the memory/catalyst no-go theorem has a genuine proof gap around CPWP instruments and the abstract overstates the claim.","tokens_in":21409,"tokens_out":9586,"would_cite":true,"duration_ms":100585,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P45","81P68"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that for pairs of states with positive discrete Wigner functions, neither finite-dimensional quantum memories nor quantum catalysts can improve the optimal discrimination success probability of classically simulable measur","keywords":["quantum state discrimination","discrete Wigner functions","classically simulable measurements","magic resource theory","channel simulation cost","quantum catalysts","quantum memory","no-go theorem"],"falsifier":"Find a pair of positive-Wigner states and an explicit finite-dimensional adaptive CPWP protocol for which the empirical per-round success rate exceeds p_PWF_succ. The most direct check is to search for a CPWP instrument that, conditioned on some outcome, maps a PWF memory state to a state with negative Wigner function; if such an instrument exists, the supermartingale step (and hence Theorem 3) fails, and a single two-round simulation would show a violation.","tokens_in":20467,"feed_emoji":"⚛️","tokens_out":6911,"duration_ms":65429,"temperature":0.7,"pith_summary":"Quantum state discrimination is usually analyzed under unrestricted measurements, but the classically simulable measurements studied here—POVMs with positive discrete Wigner functions (CSMs)—are strictly weaker. The paper asks how much that weakness can be repaired, and tests three routes: burning magic states, reusing a catalyst, or keeping an adaptive quantum memory. It shows that consumed magic genuinely helps, quantifies how much magic is needed to simulate a target non-CSM measurement channel, and exhibits a qutrit example where one Strange state raises the success probability from 2/3 to 7/9. Its principal negative result is a no-go theorem: for any pair of positive-Wigner states, the asymptotic per-round success rate of finite-dimensional memory-assisted or catalyst-assisted CPWP protocols equals the single-shot CSM value. The upshot is a clean split—consumable resources can help, reusable ones cannot, at least for this class of discrimination tasks.","feed_headline":"Quantum memory cannot improve classically simulable measurements","feed_subtitle":"Consumable magic can boost discrimination; catalysts and finite-dimensional memories cannot for Wigner-positive pairs.","key_machinery":"The discrete Wigner function and the set of completely positive Wigner-preserving (CPWP) channels carry the argument. Theorem 1 identifies PWF POVMs with CPWP measurement channels, so 'improve a CSM' becomes 'simulate a non-CPWP channel with CPWP operations and magic states'; Theorem 2 bounds that simulation cost between the mana ratio M(M)/M(ω) and a Wigner-effect existence condition. For the no-go theorem, the mechanism is the equality ||Δ⊗γ||_{PWF}=||Δ||_{PWF} for PWF γ (Lemma 1), which forces each conditional round's success probability to be at most p*, making the excess-correct-guess process a supermartingale; Azuma's inequality then converts any supposedly improved rate into a violati","core_discovery":"The central claim is that for discriminating a pair of states with positive Wigner functions, an adaptive protocol with a finite-dimensional quantum memory cannot beat the best single-shot classically simulable measurement; the same holds for catalysts, which are memories that stay fixed. The authors establish this by proving that PWF POVMs are exactly the measurement channels whose associated channel is completely positive Wigner-preserving (CPWP), which turns the question into one of channel simulation. They bound the number of magic copies needed to simulate a target channel using CPWP free operations, and give a qutrit channel (built from the Strange state) where the bounds coincide and","pith_inferences":["If the no-go theorem's component-wise PWF-preservation premise fails for some CPWP instrument, the result may be special to instruments that are Wigner-preserving in that strict sense; a testable extension is to characterize instruments that violate the premise and check whether adaptive protocols using them can beat p*.","The contrast with entanglement-assisted local discrimination suggests a possible general principle: in resource theories whose free operations are closed under conditioning, reusable catalysts and memories are inert for binary discrimination of free states; testing this in other resource theories would show how far the result generalizes.","The paper leaves open whether catalysts or memories help for non-PWF states; a numerical route is to run the SDP on a pair of states with small Wigner negativity and check whether the optimal success rate grows with memory dimension.","The tight simulation-cost example hints that the Strange-state measurement channel is a minimal non-CSM measurement; searching for other minimal non-PWF effects could yield a small catalog of elementary magic-consuming measurements."],"forward_implications":["Any finite-dimensional reusable memory or catalyst can be ignored when discriminating positive-Wigner state pairs: the best achievable rate is exactly the single-shot CSM success probability.","Consumable magic resources can strictly raise CSM discrimination power, as shown by the Strange-state example, while reusable finite-dimensional resources cannot.","The channel-simulation bounds give a quantitative answer to how much magic is needed to implement a measurement beyond CSMs, and Proposition 1 provides a case where the answer is exactly one Strange state.","The SDP formulation and its dual provide a numerical method to compute magic-assisted success probabilities for a fixed number of resource copies, enabling estimates of the magic cost of matching unrestricted measurements.","The results draw a clear line in restricted-measurement state discrimination: improvements from quantum resources must come from consumption, not reuse, at least for Wigner-positive input states."],"fun_headline_variants":["Quantum memories and catalysts don't boost simulable measurements","Magic improves simulable measurements; memories and catalysts don't","No-go: finite memories and catalysts fail for Wigner-positive pairs","Consumable magic boosts CSMs, but catalysts and memories can't"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof of the no-go theorem assumes that every adaptive CPWP instrument preserves positive Wigner functions component-wise, so the memory state after conditioning on any history γC(F_j) is always PWF; the paper asserts this without a formal definition, and without it the per-round bound Pr(X_{j+1}=1|F_j) ≤ p* does not follow.","fun_headline_variants_meta":{"raw":{"variants":["Quantum memories and catalysts don't boost simulable measurements","Magic improves simulable measurements; memories and catalysts don't","No-go: finite memories and catalysts fail for Wigner-positive pairs","Consumable magic boosts CSMs, but catalysts and memories can't"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000299,"raw_usage":{"total_tokens":1554,"prompt_tokens":719,"completion_tokens":835,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":463,"completion_tokens_details":{"reasoning_tokens":764}},"tokens_in":463,"tokens_out":835,"duration_ms":8090,"temperature":1.0,"reasoning_tokens":764,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T13:33:03.463107+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a pair of positive-Wigner states and an explicit finite-dimensional adaptive CPWP protocol for which the empirical per-round success rate exceeds p_PWF_succ. The most direct check is to search for a CPWP instrument that, conditioned on some outcome, maps a PWF memory state to a state with negative Wigner function; if such an instrument exists, the supermartingale step (and hence Theorem 3) fails, and a single two-round simulation would show a violation.","supporting_citations":[],"review_version":1}