{"id":"b0d8cded-9165-46e3-9fa7-bcc80910e5e0","arxiv_id":"2411.16504","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Alice's and Bob's causal perspectives in the quantum switch cannot be related by any fixed change of subsystem decomposition, so they are not equivalent descriptions of the same evolution.","lead":"The paper formalises how quantum circuits, including cyclic ones, can be rewritten with different choices of which subsystems carry the information. It proves that Alice's and Bob's causal perspectives in the quantum switch cannot be converted into each other by any such subsystem change, so the two perspectives are not equivalent descriptions of the same process.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the no-go proof in Appendix D is internally consistent and the fixed-unitary, discrete-circuit scoping is explicitly stated.","rationale":"The paper formalizes subsystem decompositions of quantum circuits and proves that Alice's and Bob's causal perspectives in the quantum switch cannot be interconverted by a fixed unitary change of tensor factorization. I checked the central no-go argument in Appendix D. The strategy is sound: assuming a single J satisfying Eq. (11) for all operations, one applies it to the two particular operation choices (U_B = sigma_X and U_B = sigma_Y) and sums. The resulting operators satisfy J Omega^(A) J^dagger = Omega^(B). Since Tr[Omega Omega^dagger] is invariant under unitary similarity, the two Frobenius norms must match. A direct tensor-product trace computation confirms the claimed values: each single K_temp has Tr[K K^dagger] = 2^14; for Alice's perspective the cross term Tr[K_A(sigma_X) K_A(sigma_Y)^dagger] gets a factor 2 from each of the two controlled-U_B steps and a factor 4 from each of the five identity/unitary steps, giving 2^12, hence Tr[Omega^(A) Omega^(A)^dagger] = 2^15 + 2^13; for Bob's perspective the single U_B step yields Tr[sigma_X sigma_Y^dagger] = 0, so the cross term vanishes and the trace is 2^15. The contradiction is genuine. The only soft point is definitional: the no-go holds for the paper's notion of a subsystem decomposition as a single fixed unitary J on the 16-qubit temporal Hilbert space of a discrete circuit. The authors explicitly restrict their conclusion to this setting and point to continuous-time formulations as an open question. Thus the claim is appropriately scoped and the proof is internally consistent; no load-bearing technical objection remains.","tokens_in":14302,"tokens_out":17364,"duration_ms":164869,"concrete_test":"Independently recompute the two Frobenius norms in Appendix D using explicit matrix representations or tensor-network contraction for the operators in Eqs. (15) and (21) with U_B = sigma_X and U_B = sigma_Y, U_A = 1, |psi> = |00>, <phi| = <00|. Verify Tr[Omega^(A) Omega^(A)^dagger] = 2^15 + 2^13 and Tr[Omega^(B) Omega^(B)^dagger] = 2^15; if either value shifts, the no-go proof would need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I find no load-bearing technical concern. The central no-go Eq. (11) is proven by a valid invariant argument: if a fixed isomorphism J existed for all operations, it would also relate the summed operators Omega^(A), Omega^(B) defined in Eqs. (26)-(27), and would therefore preserve Tr[Omega Omega^dagger]. The Appendix D trace values (2^15+2^13 vs 2^15) are reproduced by a direct tensor-product contraction: each single K_temp has Tr[K K^dagger]=2^14; Alice's perspective has a cross term from the two controlled-U_B steps giving 2^12, while Bob's perspective has a vanishing cross term because Tr[sigma_X sigma_Y^dagger]=0 at its single U_B step. The contradiction is genuine. The only soft point is scope: the result applies to the authors' formalization of a subsystem decomposition as a single fixed unitary J on the 16-qubit temporal Hilbert space of a discrete circuit. This caveat is openly acknowledged in the discussion ('open question whether, in a continuous framework, such a transformation between causal perspectives might be possible'), so it does not undermine the stated conclusion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a Hilbert-space-level description of quantum circuits, including cyclic ones, in terms of 'circuit operators' acting on the global Hilbert space of all temporal systems, and formalizes a change of subsystem decomposition as conjugation by a fixed unitary J (Eqs. (4)-(5)). It applies this framework to the quantum switch: Alice's and Bob's causal perspectives are each embedded into the cyclic process-matrix description via extended Hilbert spaces (Eqs. (9)-(10), Appendices B-C). The central result is Eq. (11): there is no fixed unitary J mapping Alice's temporal circuit operator to Bob's for arbitrary preparations, unitaries, and measurements. Appendix D proves this by summing the two circuit operators for UB = sigma_X and UB = sigma_Y and comparing Tr[Omega Omega-dagger], which evaluates to 2^15+2^13 for Alice's perspective and 2^15 for Bob's. The authors explicitly scope the result to the discrete-circuit setting and leave the continuous-time framework as an open question.","tokens_in":14487,"tokens_out":12438,"duration_ms":114581,"significance":"The result is a substantive negative statement: under the natural discrete fixed-isomorphism formalization, the two causal perspectives in the quantum switch cannot be regarded as different subsystem decompositions of the same time-delocalised circuit. The proof in Appendix D is explicit and self-contained, uses a correct unitary invariant, and does not rely on free parameters or hidden assumptions. The paper also gives a clear formal framework for subsystem decompositions of quantum circuits that is likely to be useful beyond this particular no-go result. I find no load-bearing technical gap; the discrete-time, fixed-J scope is acknowledged openly in the Discussion, so the stress-test concern about scope does not amount to a defect.","major_comments":[],"minor_comments":[{"comment":"The right-hand side of Eq. (23) labels the map as JA, but this equation defines Bob's isomorphism JB; the label should be corrected.","section":"Appendix C, Eq. (23)"},{"comment":"The trace values Tr[Omega(A) Omega(A)^dagger] = 2^15 + 2^13 and Tr[Omega(B) Omega(B)^dagger] = 2^15 are quoted without the intermediate contraction steps; adding those steps would make the self-contained proof easier to verify.","section":"Appendix D"},{"comment":"The claim that the two circuit operators are unitarily similar for a fixed choice of operations is explained qualitatively but not constructed explicitly; a short explicit J would remove a small presentation gap.","section":"Appendix D, first paragraph"},{"comment":"The displayed formula for C_{S_i}[M] is dense and the index contractions are not spelled out; a brief explanation of how the double-ket bra and ket implement the partial trace would improve readability.","section":"Eq. (2)"}],"recommendation":"accept","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this paper's main result is a no-go theorem, Eq. (11). The two causal perspectives in the quantum switch, each realized by a temporal circuit with a time-delocalized subsystem description, cannot be mapped into each other by any fixed unitary isomorphism J on the global Hilbert space. That is a new impossibility result, and the proof in Appendix D is explicit and checkable.\n\nThe framework itself is partly a reformulation: circuit superoperators and changes of subsystem decomposition are close to quantum-comb and link-product ideas, and the authors acknowledge the prior time-delocalised-subsystem work. But the impossibility theorem is genuinely new. The proof strategy is good: for any fixed choice of operations the two circuit operators are unitarily similar, so the authors sum two carefully chosen choices (UB = sigma_X and sigma_Y) and compare Tr[Omega Omega^dagger]. The trace values are 2^15 + 2^13 for Alice's perspective and 2^15 for Bob's. I checked the contraction; the difference comes from a cross term that survives for Alice but vanishes for Bob because Tr[sigma_X sigma_Y^dagger] = 0. The contradiction is real.\n\nThe paper is honestly scoped. The no-go holds for the authors' definition of a subsystem change as a single fixed isomorphism on the discrete temporal Hilbert space. They explicitly say that continuous-time or operation-dependent transformations remain open, so the result does not overclaim. The appendices are detailed enough to verify the two isomorphisms JA and JB and the reduction to the process matrix.\n\nSoft spots are minor. The trace values in Appendix D are quoted without the step-by-step arithmetic, so a referee will want to see the line-by-line computation, but it is straightforward. There is a small typo in Appendix C: Eq. (23) defines JB but the text says JA. Nothing load-bearing.\n\nWho is this for? People working on indefinite causal order, causal reference frames, and time-delocalised subsystems. It settles a question that has been floating around the literature: whether the two causal perspectives in the switch are just different subsystem descriptions of the same evolution. They are not, under the discrete fixed-isomorphism notion. That is worth publishing.\n\nSerious referee: yes. I would send this to peer review rather than desk reject. I would also cite it in my own work on causal perspectives.","headline":"A clean, checkable no-go result: Alice's and Bob's causal perspectives in the quantum switch are not related by any fixed subsystem-decomposition isomorphism, and the proof is worth taking seriously.","tokens_in":15004,"tokens_out":1166,"would_cite":true,"duration_ms":12651,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P68"],"pacs":[],"model":"deepseek-v4-flash","headline":"The two causal perspectives of the quantum switch cannot be related by any change of subsystem decomposition, so they are not two equivalent descriptions of the same process.","keywords":["indefinite causal order","quantum switch","time-delocalised subsystems","causal perspectives","subsystem decompositions","circuit operator","quantum reference frames","causal reference frames"],"falsifier":"The decisive check is the trace calculation in Appendix D: verify whether ${\\rm Tr}[\\Omega^{(A)}\\Omega^{(A)\\dagger}] = 2^{15}+2^{13}$ and ${\\rm Tr}[\\Omega^{(B)}\\Omega^{(B)\\dagger}] = 2^{15}$ for the two summed operators defined there. If those numbers are equal, or if a single unitary $J$ satisfying Eq. (11) for all operations can be written down explicitly, the no-go claim is wrong.","tokens_in":14088,"feed_emoji":"⚛️","tokens_out":11817,"duration_ms":97140,"temperature":0.7,"pith_summary":"This paper asks whether the two causal perspectives of the quantum switch—Alice's view and Bob's view of a process with indefinite causal order—are just two ways of carving up the same underlying quantum evolution. The authors formalize a subsystem decomposition of a quantum circuit by unfolding the circuit into a single operator on the global Hilbert space of all systems at all times, and show how that operator changes under an arbitrary unitary relabelling of the subsystems. Their main result is negative: for the quantum switch there is no such relabelling that turns Alice's temporal circuit into Bob's, for arbitrary choices of the local operations. This matters because the intuition that causal perspectives should be equivalent underlies recent relational and quantum-reference-frame pictures of indefinite causal order. In the discrete circuit setting where such processes are usually studied, distinct causal perspectives are not equivalent descriptions of the same evolution.","feed_headline":"No subsystem change links the quantum switch's two causal perspectives","feed_subtitle":"Alice's and Bob's views of a causally indefinite process are provably inequivalent in discrete circuits.","key_machinery":"The load-bearing object is the circuit operator: the tensor product of the Kraus operator at every time step of a circuit, acting on the global Hilbert space formed by all systems at all times. A change of subsystem decomposition is defined as a unitary isomorphism $J$ on that global space, under which the operator transforms by conjugation, $K \\mapsto J K J^\\dagger$ (Eqs. (4)–(5)). Because conjugation preserves ${\\rm Tr}[\\Omega\\Omega^\\dagger]$, comparing this invariant for sums of circuit operators built from two specific choices of Bob's unitary is what proves the no-go result. The two causal perspectives under study are the temporal circuits of Fig. 4, each with a target and a control qubit at eight time steps; each is separately related to the cyclic quantum-switch circuit by an isomorphism $J_A$ or $J_B$ that introduces extra ancilla systems $E_A$ or $E_B$ which must be traced out.","core_discovery":"The central claim is Eq. (11): there is no unitary isomorphism $J$ from the 16-qubit temporal Hilbert space of Alice's perspective to that of Bob's such that $J K_{\\rm temp}^{(A)}(|\\psi\\rangle,U_A,U_B,\\langle\\phi|) J^{\\dagger} = K_{\\rm temp}^{(B)}(|\\psi\\rangle,U_A,U_B,\\langle\\phi|)$ for all preparations $|\\psi\\rangle$, unitaries $U_A,U_B$, and measurements $\\langle\\phi|$. For any single fixed choice of operations the two circuit operators are unitarily similar, so a superficial check can suggest equivalence; the obstruction appears when multiple choices of operations are combined. Taking $U_A = \\mathbb{1}$, preparation and measurement $|00\\rangle$ and $\\langle 00|$, and $U_B$ equal to $\\sigma_X$ or $\\sigma_Y$, the paper evaluates ${\\rm Tr}[\\Omega^{(A)}\\Omega^{(A)\\dagger}] = 2^{15}+2^{13}$ and ${\\rm Tr}[\\Omega^{(B)}\\Omega^{(B)\\dagger}] = 2^{15}$, a difference that conjugation by any fixed $J$ would preserve. Hence the two causal perspectives are incompatible subsystem decompositions of the same evolution in the discrete setting.","pith_inferences":["If the no-go survives in a continuous-time or gravitational setting, then relations between causal perspectives would have to be operation-dependent maps rather than fixed subsystem relabellings, which would require a new notion of coordinate transformation.","The trace-invariant comparison could be turned into a general equivalence test: given two proposed time-delocalised realisations of the same process, compare ${\\rm Tr}[\\Omega\\Omega^\\dagger]$ over a small set of operation choices instead of attempting to construct an isomorphism.","A structural conjecture suggested by the proof is that any two causal perspectives whose constructions require different auxiliary systems that must be traced out will be inequivalent in this sense; the quantum switch would then be one instance of a general pattern.","The same method may apply to other causally indefinite processes with multiple party perspectives, so the result is not necessarily specific to the quantum switch."],"forward_implications":["Alice's and Bob's causal perspectives in the quantum switch cannot both be subsystem decompositions of one and the same global evolution, so the process-matrix description does not act as an observer-neutral reference from which either perspective can be reached by a fixed subsystem relabelling.","The probability of a circuit is invariant under any change of subsystem decomposition, so the framework supplies a consistency check that any candidate equivalence between two circuit descriptions must pass.","The impossibility holds for the whole family of possible operations; for any single fixed choice of all operations the two circuit operators are unitarily similar, which means the no-go is not visible in a single run of the switch.","The formalism also covers cyclic and consistent circuits, giving a common language for time-delocalised realisations of indefinite causal order beyond the quantum switch.","Whether a continuous-time formulation could restore equivalence between causal perspectives remains open, and the authors identify it as the key question for hypothetical gravitational realisations of indefinite causal order."],"supporting_citations":[{"why":"Supplies the process matrix framework and the generalised Born rule that define probabilities for processes with indefinite causal order, including the quantum switch.","marker":"[1]"},{"why":"Introduces the quantum switch as the canonical example of a causally indefinite process whose perspectives are studied here.","marker":"[6]"},{"why":"Shows that certain causally indefinite processes can be realised as standard temporal evolutions on time-delocalised subsystems, the notion formalised in this paper.","marker":"[17]"},{"why":"Extends time-delocalised realisations to arbitrary processes and causal inequalities, providing the general context for the subsystem-decomposition framework.","marker":"[18]"},{"why":"Introduces the notion of causal perspectives (causal reference frames) for the quantum switch that the paper analyses.","marker":"[25]"},{"why":"Argues for causal perspectives from quantum reference frames in a continuous setting, giving the intuitive expectation of equivalence that the no-go result challenges.","marker":"[26]"},{"why":"Provides the detailed account of the two causal perspectives in the quantum switch that yields the temporal circuits of Fig. 4.","marker":"[27]"},{"why":"Gives the process matrix description of the quantum switch that underlies the cyclic circuit and the isomorphisms in Appendices B and C.","marker":"[32]"}],"fun_headline_variants":["Quantum switch: Alice's and Bob's views are provably inequivalent","Two causal perspectives of a quantum switch can't be unified","Alice and Bob's views of the quantum switch: no common description","Quantum switch's causal perspectives aren't equivalent descriptions","No single change of subsystems equates two quantum switch views"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes that a change of subsystem decomposition is one fixed unitary relabelling of the entire collection of quantum systems across time, applied identically for every choice of the preparation, the unitaries, and the final measurement, and that the two discrete circuits of Fig. 4 correctly represent the two causal perspectives; continuous-time descriptions could behave differently.","fun_headline_variants_meta":{"raw":{"variants":["Quantum switch: Alice's and Bob's views are provably inequivalent","Two causal perspectives of a quantum switch can't be unified","Alice and Bob's views of the quantum switch: no common description","Quantum switch's causal perspectives aren't equivalent descriptions","No single change of subsystems equates two quantum switch views"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000685,"raw_usage":{"total_tokens":3143,"prompt_tokens":1016,"completion_tokens":2127,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":632,"completion_tokens_details":{"reasoning_tokens":2043}},"tokens_in":632,"tokens_out":2127,"duration_ms":14768,"temperature":1.0,"reasoning_tokens":2043,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T13:02:41.817071+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The decisive check is the trace calculation in Appendix D: verify whether ${\\rm Tr}[\\Omega^{(A)}\\Omega^{(A)\\dagger}] = 2^{15}+2^{13}$ and ${\\rm Tr}[\\Omega^{(B)}\\Omega^{(B)\\dagger}] = 2^{15}$ for the two summed operators defined there. If those numbers are equal, or if a single unitary $J$ satisfying Eq. (11) for all operations can be written down explicitly, the no-go claim is wrong.","supporting_citations":[{"cited_title":"Composition rules for quantum processes: a no-go theorem","cited_arxiv_id":"1806.10374","evidence_quote":"Introduces the notion of causal perspectives (causal reference frames) for the quantum switch that the paper analyses."},{"cited_title":"Noncausal Page-Wootters circuits","cited_arxiv_id":"2105.02304","evidence_quote":"Provides the detailed account of the two causal perspectives in the quantum switch that yields the temporal circuits of Fig. 4."}],"review_version":1}