{"id":"6db749ff-cfd1-4088-9390-e360a3a802df","arxiv_id":"2509.10430","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Locally implementable product unitaries can be globally distinguishable but LOCC-indistinguishable, and vice versa, depending on whether probing states are adaptive or fixed and whether probes are separable or entangled.","lead":"The paper compares how well a single unknown quantum operation can be identified when the whole system is measured jointly versus when only local measurements and classical messages are allowed. It reports three examples where these two capabilities differ, exposing limits on what separated laboratories can do even when the operation is made of independent parts.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 2's LOCC-indistinguishability depends on excluding shared entangled probes, yet the abstract and title present the result under unqualified 'LOCC'; under the common channel-discrimination convention allowing an entangled input probe, the pair becomes LOCC-distinguishable.","rationale":"Agree with the reader's weakest-assumption identification. The concern is not a mathematical error inside the paper's stated model; the proofs are explicit and the examples are checkable. Rather, the most load-bearing issue is that the headline separation (i) is stated under unqualified 'LOCC' while the formal model silently restricts the probe to be a tensor product of locally prepared states. That restriction is what makes Theorem 2's impossibility true. If the common reading in which an entangled input probe is allowed is adopted, the theorem's conclusion fails; if the paper's reading is retained, the abstract and theorem statements need an explicit qualifier such as 'LOCC with product-prepared probes.' The reader's conditional verdict is appropriate: the correct fix is a definitional and scope revision, not a change of the internal calculations. I do not see an internal inconsistency from Theorem 4, whose Bell probes are local ancillas rather than a shared cross-party state; the product-probe restriction is violated only by the global side of Theorem 2 and by any external reading of 'LOCC' that permits pre-shared entanglement. Secondary issues—the unproved Observation 2, reliance on the self-cited [28] for Theorem 1, and the unverified SDP claim in Proposition 1—are real but less central than the probe-model ambiguity, because they affect auxiliary or support results rather than the basic meaning of the LOCC separation.","tokens_in":10791,"tokens_out":28705,"duration_ms":259811,"concrete_test":"Use the pair (2) with parameters satisfying α+β+γ+δ=π. Supply |φ+⟩ = (|00⟩+|11⟩)/√2 as the initial probe, apply U1 or U2, and then perform the LOCC measurement that perfectly discriminates the two orthogonal output states (e.g., Walgate et al.'s protocol; here the states differ only by relative phase and are orthogonal). If this protocol is allowed, Theorem 2's LOCC-indistinguishability is contradicted. To decide whether the paper's model allows it, check whether |φ+⟩ can be written as ρ_A1B1 ⊗ ρ_A2B2 under the LDR/LDA definition: it cannot, because the reduced state on the two unitary subsystems is entangled. This one check settles that the theorem's 'LOCC' is the restricted product-probe model, not the standard LOCC-with-shared-entanglement model.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's formal LOCC model ('Protocols for LOCC distinguishability') has each party k independently prepare ρ_AkBk, so the collective initial state is a tensor product across the spatially separated systems. This excludes the shared Bell probe |φ+⟩ used in the global protocol of Theorem 2 (Appendix A). The exclusion is load-bearing: if an LOCC protocol were allowed to begin with that same |φ+⟩, then after applying U1 or U2 from (2) the two output states are orthogonal, and any two orthogonal pure states are perfectly distinguishable by LOCC (Walgate et al., PRL 85, 4972). Hence the claimed 'not LOCC distinguishable' conclusion in Theorem 2 holds only for product-prepared probes, not for LOCC with pre-shared entanglement. The formal section does state the product-probe restriction, but the abstract and title say 'under LOCC' without this caveat, and the theorem statements repeat the unqualified phrase. If a reader takes 'LOCC' in the channel-discrimination sense where the input probe is part of the problem specification and may be entangled, Theorem 2 is false. The central separation (i) is therefore an artifact of a nonstandard, incompletely advertised restriction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies single-shot discrimination of locally implementable (product) multipartite unitaries, comparing global operations with LOCC and classifying strategies as restricted (fixed probing states) or adaptive (probing states chosen based on previous outcomes). The main results are: Theorem 1 (for bipartite qubit unitaries, adaptive and restricted LOCC are equivalent), Theorem 2 (a family of product qubit unitaries is globally distinguishable with restricted strategies but not LOCC-distinguishable), Theorem 3 (a set of four product qutrit unitaries is LOCC-distinguishable adaptively but globally indistinguishable with restricted strategies), Proposition 1 (asymmetry of LOCC distinguishability for the set in Theorem 3), and Theorem 4 (a set of five product qubit unitaries is globally indistinguishable with single-system probes but LOCC-distinguishable with an entangled probe). The proofs are mainly explicit constructions in the appendices, with orthogonality conditions checked directly, and the paper also proves structural equivalences among strategy classes.","tokens_in":11061,"tokens_out":10997,"duration_ms":98765,"significance":"If the results are read under a clearly and consistently stated probe model, the paper offers concrete examples of global/LOCC separations for locally implementable unitaries, a relatively underexplored question. The constructions in Appendices A and B are explicit and self-contained, and the paper usefully distinguishes restricted vs adaptive strategies. However, the significance is substantially undercut by the manuscript's inconsistent treatment of whether LOCC protocols may begin with an entangled probe shared across the parties: Theorem 2 is false under the standard convention allowing an arbitrary input probe, while Theorem 4's LOCC protocol uses exactly such a shared entangled probe. These issues must be resolved before the claimed separations can be accepted.","major_comments":[{"comment":"The LOCC model defined in this section restricts each party to independently prepare its own probe state ρ_AkBk, so the collective input is a tensor product across the parties and no pre-shared entanglement is available. This restriction is not stated in the abstract, the introduction, or the theorem statements, which use the unqualified phrase 'under LOCC'. Under the standard channel-discrimination convention in which the input probe is part of the strategy and may be entangled across the parties, Theorem 2 is false: the shared Bell state |φ+> used in the GDR proof in Appendix A is an allowed input, and after applying U1 or U2 from (2) the output states are orthogonal, so they are perfectly distinguishable by LOCC. The claimed separation (i) therefore holds only for the paper's restricted probe model, and the paper must state this restriction in the abstract and in every theorem statement.","section":"Protocols for LOCC distinguishability; Theorem 2"},{"comment":"Theorem 4's LOCC protocol begins with the shared Bell state |φ+> as the initial probing state, which is not a product of independently prepared local states and is therefore not allowed by the LOCC model defined in the section 'Protocols for LOCC distinguishability'. Conversely, if the authors intend to allow entangled probes under LOCC, then Theorem 2's LOCC-indistinguishability claim fails. The manuscript therefore uses two incompatible notions of LOCC in Theorems 2 and 4. The authors should adopt one convention and revise the affected statements and proofs accordingly.","section":"Theorem 4 and Appendix D"},{"comment":"The proof that the unitaries in (4) are globally indistinguishable with single-system probes is not given in full. It asserts that the only triples Alice can eliminate are (X,X,XZ) and that all other listed triples are impossible, but no argument is provided to rule out other measurements on a single-qubit probe. Since this is the basis for the first half of Theorem 4, a complete proof is needed.","section":"Appendix D, first paragraph"}],"minor_comments":[{"comment":"The phrases 'under LOCC' and 'indistinguishable under LOCC' should be qualified as 'LOCC with independently prepared (product) probes' once the model in Section 'Protocols for LOCC distinguishability' is adopted.","section":"Abstract and Introduction"},{"comment":"The ordering diagram GDR≤GDA≤≤LDR≤LDA is visually confusing; the intended nesting of the two chains (global vs LOCC) should be made explicit with labeled arrows or a clear sentence.","section":"Observation 1, Eq. (1)"},{"comment":"In the GDR proof, the phrase 'from (1)' is misleading; the distinguishability follows from orthogonality of the evolved states, not from the ordering diagram in Observation 1.","section":"Appendix A"},{"comment":"The notation '1(R1)+e^{i2π/3}(R2)+e^{i4π/3}(R3)' is hard to read; it should be written with explicit multiplication, e.g., R1 + e^{i2π/3}R2 + e^{i4π/3}R3. Also, 'Krauss operators' should be 'Kraus operators'.","section":"Appendix B"},{"comment":"The text lists Bob's distinguishable pairs as including '(H,H)', but from (4) the relevant pair is (H, \\bar H); this appears to be a typo and should be corrected.","section":"Appendix D"},{"comment":"The proof relies on Theorem 4 of the authors' own preprint arXiv:2504.14499 as a black box; this dependency should be stated explicitly in the main text, and the status of that preprint should be clarified.","section":"Theorem 1 proof"},{"comment":"The see-saw SDP method is described in a single sentence; to make Proposition 1 verifiable, the authors should give the full SDP formulation or a reference that completely specifies the method.","section":"Appendix C"}],"recommendation":"major_revision","confidential_remarks":"The core mathematical constructions in the appendices appear correct for the paper's explicit probe restrictions, but the inconsistent definition of LOCC between Theorem 2 and Theorem 4 is a fundamental issue that the authors must resolve. I also note that Theorem 1 depends on an unpublished self-cited preprint (arXiv:2504.14499); the editor may wish to check whether that result has been accepted or refereed elsewhere."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the useful core: the paper asks a good question and gives explicit constructions for a useful taxonomy. Theorems 3 and 4 check out as separations within the model where each party can use ancillas but the initial state across the parties is product. Theorem 3's proof is a real orthogonality argument, and Theorem 4's local protocol uses local Bell states between main and ancilla, so it is not the victim of the shared-entanglement objection. The examples are checkable.\n\nThe soft spot is the headline: 'LOCC' is used in a nonstandard sense. In the protocol section, each party independently prepares rho_AkBk, so the collective probe is a product across parties. No initial entanglement between Alice and Bob is allowed. The abstract, title, and theorem statements do not say this. For Theorem 2 that restriction is load-bearing: the global protocol uses a shared Bell probe across parties, the output states are orthogonal, and orthogonal pure states are LOCC-distinguishable. So if 'LOCC' is read in the channel-discrimination sense—choose any input probe, then act by LOCC—Theorem 2 is false. The paper needs to either rename the model (e.g., 'LOCC with product-prepared probes') throughout or restrict the claims accordingly. This is not cosmetic; it changes the meaning of the central result.\n\nOther issues are smaller but real. Observation 2, used in Proposition 1, is asserted without proof. Proposition 1's SDP calculation is described as see-saw, with no code, no certificate, and no rigorous argument; as written it is an unverified numerical claim. Theorem 1 leans on an unpublished self-cited preprint—maybe fine, but a referee should verify the dependency.\n\nVerdict: the paper deserves a serious referee, but the authors should be asked to fix the model terminology and prove or remove the auxiliary steps before publication. If they fix those, the remaining examples are a legitimate contribution to the LOCC-versus-global channel-discrimination story. For now, I would not cite the LOCC claims as stated.","headline":"Real examples inside a nonstandard probe model, but the 'LOCC' label hides a load-bearing restriction that makes Theorem 2 false under the standard channel-discrimination convention.","tokens_in":11580,"tokens_out":5888,"would_cite":false,"duration_ms":412375,"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":"This paper establishes that for locally implementable bipartite unitaries, global and LOCC distinguishability can separate in both directions: some pairs are globally but not locally distinguishable, while other sets are locally but not…","keywords":["unitary discrimination","LOCC","product unitaries","adaptive strategies","restricted strategies","quantum channel discrimination","nonlocality without entanglement","entangled probes"],"falsifier":"For the qutrit set (3), numerically maximize over all four-qutrit probing states $|\\psi\\rangle$ the quantity $\\min_{i\\neq j} |\\langle\\psi| V_i^\\dagger V_j \\otimes 1 \\otimes 1 |\\psi\\rangle|$; if any probe makes all pairs orthogonal, the claimed global restricted indistinguishability of Theorem 3 collapses.","tokens_in":10545,"feed_emoji":"⚛️","tokens_out":8742,"duration_ms":69516,"temperature":0.7,"pith_summary":"The paper studies single-shot discrimination of locally implementable (product) multipartite unitaries under global operations versus LOCC. It classifies strategies into restricted (fixed probing states) and adaptive (probing states chosen based on earlier outcomes) and asks whether the two paradigms ever disagree. The answer is yes, in both directions: the authors exhibit pairs of qubit unitaries that a global user can distinguish with a fixed probe but that LOCC cannot distinguish even adaptively; four-qutrit sets that LOCC can distinguish adaptively but no global fixed-probe strategy can; and five-qubit unitaries that are globally indistinguishable to any adaptive strategy using separable probes yet LOCC-distinguishable with an entangled probe. The results also show that for bipartite qubit unitaries adaptive and restricted LOCC are equivalent, and that LOCC distinguishability can depend on which party starts.","feed_headline":"Global and LOCC discrimination split in both directions","feed_subtitle":"Three explicit unitary sets show global access beats LOCC in one case, LOCC beats global in two others.","key_machinery":"The carrying mechanism is a four-way classification of discrimination protocols (global restricted, global adaptive, LOCC restricted, LOCC adaptive), together with the standard criterion that two unitaries $U_1, U_2$ are perfectly distinguishable iff the convex hull of the eigenvalues of $U_1^\\dagger U_2$ contains zero (the $\\min|\\mathrm{con}\\{\\cdot\\}|$ condition). The proofs hinge on constructing unitary sets with engineered phase relations---e.g. $\\alpha+\\beta+\\gamma+\\delta = \\pi$ for Theorem 2---so that a chosen probe (the Bell state $|\\phi^+\\rangle$) makes evolved states orthogonal while local probes can never achieve the eigenvalue-convex-hull condition. For Theorem 3, the asymmetry comes from the order of operation: Alice's initial qutrit measurement splits Bob's task into distinguishable pairs, whereas Bob's starting measurement cannot eliminate enough candidates.","core_discovery":"On the paper's own terms, the central discovery is a set of counterexamples to the intuition that global access is always at least as powerful as LOCC for telling apart unitary processes. Theorem 2 gives a pair of qubit unitaries of the form $U_1 = A_1\\otimes R_1$, $U_2 = A_2\\otimes R_2$ with phases satisfying $\\alpha+\\beta+\\gamma+\\delta = \\pi$; these are globally distinguishable with a restricted strategy (using the shared Bell state $|\\phi^+\\rangle$ as probe, the evolved states are orthogonal) yet are not distinguishable under LOCC, even adaptively. Theorem 3 gives four qutrit unitaries $\\{1_3\\otimes 1_3, 1_3\\otimes\\Omega, \\Omega\\otimes 1_3, \\Omega\\otimes B\\}$ that are LOCC-distinguishable when Alice starts adaptively but globally indistinguishable under any restricted strategy; moreover they are LOCC-indistinguishable if Bob starts (Proposition 1). Theorem 4 gives five qubit unitaries $\\{1_2\\otimes 1_2, Z\\otimes X, X\\otimes H, X\\otimes \\bar H, XZ\\otimes H\\}$ that are globally indistinguishable to every adaptive strategy using single-system separable probes yet LOCC-distinguishable when Alice and Bob share a Bell probe. The paper also proves Theorem 1: for bipartite qubit unitaries, adaptive and restricted LOCC strategies coincide for perfect distinguishability.","pith_inferences":["A natural next step is a multipartite example separating global adaptive strategies from both global restricted and LOCC adaptive; the paper's Observation 2 already reduces the bipartite case to a dichotomy.","These separations are sensitive to the probe-preparation model; allowing free pre-shared entanglement in LOCC may erase the Theorem 2-type gaps, so the boundary between 'LOCC with local probes' and 'LOCC with shared resources' deserves explicit treatment in future work.","The asymmetry of Proposition 1 hints that 'who speaks first' is a structural resource for unitary discrimination, analogous to asymmetric state-discrimination scenarios; one could test whether similar starter-dependence appears in other channel discrimination tasks.","The Theorem 4 result suggests a practical principle: when entanglement is cheap between parties but global access is unavailable, entanglement-assisted LOCC can outperform any separable-probe global strategy---an observation with potential relevance to distributed quantum certification."],"forward_implications":["For any bipartite qubit unitary set, an optimal perfect-distinguishing LOCC protocol can always be restricted (fixed probe): adaptivity buys nothing (Theorem 1).","Global access is not strictly stronger than LOCC for unitary discrimination: the four-qutrit set (3) is LOCC-distinguishable but globally restricted-indistinguishable.","LOCC is not always weaker either: the pairs in (2) are globally distinguishable but LOCC-indistinguishable even adaptively.","Probe entanglement can activate LOCC distinguishability: the five-qubit set (4) fails under separable probes globally but succeeds via an entangled Bell probe under LOCC.","The order in which LOCC parties act can matter: the set (3) is distinguishable if Alice starts but not if Bob does."],"supporting_citations":[{"why":"Supplies the $\\min|\\mathrm{con}\\{\\cdot\\}|$ criterion used to decide when two unitaries are perfectly distinguishable.","marker":"[27]"},{"why":"Its Theorem 4 (common maximally entangled probes) is the key step in proving Theorem 1 and the local indistinguishability in Theorem 2.","marker":"[28]"},{"why":"Provides the rigorous characterization of LOCC that the paper adopts for its local protocol model.","marker":"[38]"},{"why":"Established the adaptive-versus-nonadaptive distinction for channel discrimination, which the paper lifts to the LOCC setting.","marker":"[23]"},{"why":"Gives earlier criteria for LOCC distinguishability of multipartite unitary operations, the problem class this paper extends to locally implementable unitaries.","marker":"[33]"}],"fun_headline_variants":["LOCC beats global at distinguishing some quantum unitaries","Local strategies outshine global in telling unitaries apart","Quantum unitaries: sometimes local is more powerful than global","Reversal: LOCC can distinguish where global cannot","Counterintuitive: LOCC outperforms global in unitary discrimination"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that in every LOCC protocol each party independently prepares its own probe state locally, so the parties share no initial entanglement; if LOCC were allowed to begin with a shared entangled state such as $|\\phi^+\\rangle$, some of the claimed separations (in particular Theorem 2) would fail, since any two orthogonal pure states are LOCC-distinguishable.","fun_headline_variants_meta":{"raw":{"variants":["LOCC beats global at distinguishing some quantum unitaries","Local strategies outshine global in telling unitaries apart","Quantum unitaries: sometimes local is more powerful than global","Reversal: LOCC can distinguish where global cannot","Counterintuitive: LOCC outperforms global in unitary discrimination"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00046,"raw_usage":{"total_tokens":2338,"prompt_tokens":1012,"completion_tokens":1326,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":628,"completion_tokens_details":{"reasoning_tokens":1247}},"tokens_in":628,"tokens_out":1326,"duration_ms":11822,"temperature":1.0,"reasoning_tokens":1247,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:56:11.109947+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For the qutrit set (3), numerically maximize over all four-qutrit probing states $|\\psi\\rangle$ the quantity $\\min_{i\\neq j} |\\langle\\psi| V_i^\\dagger V_j \\otimes 1 \\otimes 1 |\\psi\\rangle|$; if any probe makes all pairs orthogonal, the claimed global restricted indistinguishability of Theorem 3 collapses.","supporting_citations":[{"cited_title":"Everything you always wanted to know about locc (but were afraid to ask),","cited_arxiv_id":null,"evidence_quote":"Provides the rigorous characterization of LOCC that the paper adopts for its local protocol model."}],"review_version":1}