{"id":"63abfaf9-6b12-41b7-bd52-8b4fbc5d2280","arxiv_id":"2607.15947","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For qutrit maps, 1-undistillable channels composed with Schmidt-number-two channels are always entanglement breaking, and 1-undistillable maps are the largest cone with this property.","lead":"This paper proves a sharp result about when pairs of quantum channels can still generate entanglement through entanglement swapping, for three-level quantum systems. It shows that any channel whose Choi matrix is 1-undistillable, combined with any channel whose Choi matrix has Schmidt number at most two, always yields an entanglement-breaking composition.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified — the 2⊗3 Horodecki criterion applies cleanly to the filtered summands, and all other proof steps check out.","rationale":"The reader's verdict is ACCEPT with high confidence, and the weakest assumption is the Horodecki criterion applied to rank-two filtered Choi matrices. I agree that this is the most load-bearing external theorem in the proof, but after careful checking it is correctly applied: the criterion holds for positive semidefinite operators on 2⊗3 after normalization, the low-rank support of each summand is an embedded 2⊗3 system, and the sum of separable summands is separable. I also verified the less prominent steps that could have hidden trouble: Proposition 11's equivalence between 1-undistillability and PPT of rank-two filtered matrices (both left- and right-filtering), the use of the Choi correspondence in Theorem 6's converse, and the local-filter identity in Lemma 12. The only externally cited result with real weight is the Horodecki criterion, which is standard and does not constitute a correctness risk. The paper's self-flagging limitation in the Discussion is accurate but not an objection. Hence no change to the reader's verdict is warranted; a numerical spot-check would nonetheless be a cheap way to corroborate the key filter-PPT implication.","tokens_in":13601,"tokens_out":25216,"duration_ms":249220,"concrete_test":"Run a numerical sanity check for Proposition 11 and Theorem 4: sample random qutrit Choi matrices C_Φ in UND₁ (e.g., random PPT states) and random rank-2 X; verify that (X⊗I)C_Φ(X†⊗I) is PPT and then separable using a 2⊗3 separability SDP. Also sample a known 1-distillable state (e.g., a suitable Werner state) and confirm some rank-2 filter yields non-PPT. If all samples conform, the critical assumption is fully borne out.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I read the proof of Theorems 4, 6, and 9 in detail and found no load-bearing gap. The closest thing to an external assumption is, as the reader noted, the Horodecki criterion (PPT ⇒ separability) for 2⊗3 and 3⊗2 systems, used in Appendix B to declare each summand separable. This is sound: each summand is a rank-≤2 local filtering of a 1-undistillable Choi matrix, hence PPT by Proposition 11; normalization is irrelevant because the criterion scales; and the 2-dimensional subspace is just an embedded C². Proposition 11's equivalences (i)⇔(ii) and (i)⇔(iii) were checked, including the vectorization/transpose signs in the converse directions, and they hold. Theorem 6's converse correctly leverages Proposition 11(v)/(vi) together with the Choi bijection between SP₂ and S₂. Theorem 9's Lemma 12 was also verified: the key A⇒B local-filter identity is correct by index expansion, and the cone-closure arguments for S₂ and UND₁ are valid. The paper itself flags the Horodecki-criterion reliance in the Discussion, which is honest and not a defect. No circular reasoning, missing derivation, or parameter-fitting was found.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the EB-composability problem for ordered pairs of cones of completely positive maps, where (X,Y) is EB-composable if Φ∘Ψ is entanglement breaking for every Φ∈X and Ψ∈Y. In the qutrit setting, it proves that (UND_1, SP_2) and (SP_2, UND_1) are EB-composable (Theorem 4), and that UND_1 is exactly the largest cone of qutrit CP maps that is EB-composable with SP_2 in either order (Theorem 6). It then goes from the map-composition formulation to the full state-level entanglement-swapping picture: for cones closed under local filtering, separability of the maximally-entangled-outcome post-selected state implies separability of the state obtained from every positive operator on the intermediate systems (Theorem 9). Applying this to UND_1 and S_2 gives Corollary 10, a universal obstruction to terminal entanglement generation whenever one link is 1-undistillable and the other has Schmidt number at most two. The proofs are contained in Appendices B–E.","tokens_in":13933,"tokens_out":34025,"duration_ms":317981,"significance":"This is a substantial step beyond the qutrit PPT squared theorem. The paper replaces the pair (PPT, PPT) by the larger pair (UND_1, SP_2) and proves both a positive composition result and a sharp maximality statement: any qutrit map outside UND_1 is detected by some SP_2 test map in at least one order. The state-level extension via Theorem 9 is also valuable, since it shows that the maximally entangled-outcome criterion controls all selective measurements under a natural closure assumption. The proofs are complete and self-contained modulo standard facts: the Horodecki criterion for 2⊗3 and 3⊗2 systems, the Choi–Jamiołkowski correspondence, and the cited inclusion UND_1⊆SP_2 for qutrits. Proposition 11 is derived from the definition of 1-undistillability, and the key steps in Appendices B, D, and E check out. There is no parameter fitting, no numerical conjecture, and the maximality claim is a sharp, falsifiable statement. The paper itself explicitly acknowledges the reliance on the 2⊗3 Horodecki criterion in the Discussion.","major_comments":[],"minor_comments":[{"comment":"The step from Theorems 4 and 9 to Corollary 10 is quite compressed. It would help readers if the authors explicitly stated the Choi map for a given state ρ_AB (i.e., the CP map whose Choi matrix is proportional to ρ_AB) and explained how the two orderings in Corollary 10 correspond to the two orderings in Theorem 4. The step is justified by the Choi–Jamiołkowski correspondence and by scaling, but a few sentences would remove a potential source of confusion.","section":"Section III, Corollary 10"},{"comment":"In the definition of ρ'_AB=(I_A⊗X†)ρ_AB(I_A⊗X), it is easy to miss that this is a legitimate local filter of ρ_AB with L=I_A and R=X†. The proof also uses cyclicity of the partial trace over BC to move X and X† across ρ_AB⊗σ_CD; stating this explicitly would make the argument easier to verify.","section":"Appendix E, Lemma 12, (A)⇒(B)"},{"comment":"The phrase 'post-selected unnormalized state' for a general positive operator M_BC may be ambiguous, since the physical post-measurement state for a POVM element M is usually written with √M on both sides. For X=ρ_AB⊗σ_CD, the identity Tr_BC[√M X √M] = Tr_BC[X M] holds by spectral decomposition, so the formulation is correct; a brief remark would prevent confusion.","section":"Conjecture 1 and Theorem 9"},{"comment":"The decomposition ρ=Σ|ψ_i⟩⟨ψ_i| in the definition of Schmidt number should be explicitly allowed to include positive weights, or the vectors should be allowed to be unnormalized, to avoid ambiguity with the unnormalized maximally entangled state used throughout the paper.","section":"Appendix A, Definition of Schmidt number"}],"recommendation":"accept","confidential_remarks":"I found no load-bearing gap. The only external theorem that takes nontrivial weight is the Horodecki criterion for 2⊗3 and 3⊗2 systems; it is applied correctly to rank-≤2 filtered summands and is explicitly acknowledged in the Discussion. The paper is a clean, complete contribution that goes beyond the qutrit PPT squared theorem and gives a sharp maximality result. I support acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good paper to look at if you care about the PPT squared conjecture. The qutrit result here is not just another composition theorem. They define EB-composability for cones of CP maps, prove that the cone of 1-undistillable maps is exactly the largest cone EB-composable with the Schmidt-number-two cone SP2, in either order. That is a sharp boundary result, not a sufficiency statement. The second half shows that this map-level conclusion extends to arbitrary selective measurements in entanglement swapping, provided the state cones are closed under local filtering. That bridge is the operational payoff.\n\nWhat's new: Theorems 4, 6, and 9. The proofs are complete and checkable. They lean on the Horodecki criterion for 2x3 and 3x2 systems to declare filtered summands separable. That is a standard external theorem, and the paper flags it in the Discussion. I don't see a gap. Proposition 11's equivalences are proven, and I checked the vectorization signs. Corollary 10's passage from map composition to arbitrary states is terse but justified via the CJ correspondence and scaling; the stress test confirmed it.\n\nSoft spots: The result is qutrit-specific. The proof mechanism dies in higher dimensions because PPT does not imply separability on 2xd subspaces. The paper knows this and says so. Minor: the maximality is relative to the full testing cone SP2; with a smaller testing cone the compatible cone can be larger, as their Werner-Holevo example shows. That is not a flaw, but it is a bound on how general the story is. The order-asymmetry example in Appendix C is nice and relevant.\n\nCitation pattern looks fine: they use the known qutrit PPT squared result, the 1-undistillability literature, and the Schmidt number cone papers. No suspicious self-citation beyond normal. No fitted parameters, no circularity.\n\nWho this is for: anyone working on entanglement swapping, distillability, or the PPT squared problem. The framework of EB-composability is likely to be reused. I'd take it seriously as a paper in a good journal; it's not a breakthrough but it is a clean, correct, meaningful step beyond the PPT squared conjecture.\n\nRecommendation: send it to peer review. If the referee agrees with the proofs, accept with minor revisions; the main thing is the terse step in Corollary 10 could be expanded.","headline":"A clean, correct qutrit result that sharpens the PPT-squared problem and gives a maximality characterization; worth a serious referee.","tokens_in":14353,"tokens_out":1408,"would_cite":true,"duration_ms":15629,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P40","81P45"],"pacs":[],"model":"deepseek-v4-flash","headline":"For qutrit maps, composing any 1-undistillable map with any Schmidt-number-two map always yields an entanglement breaking map, in either order, and the 1-undistillable class is exactly the largest such cone.","keywords":["Entanglement swapping","PPT squared conjecture","1-undistillable states","Schmidt number","Entanglement breaking maps","Qutrit completely positive maps","EB-composability","Choi-Jamiołkowski isomorphism"],"falsifier":"Find a qutrit completely positive map with a 1-distillable Choi matrix for which composition with every Schmidt-number-two map, in both orders, is still entanglement breaking; Theorem 6 says such a map cannot exist. Alternatively, exhibit a two-qutrit state in UND_1 and a two-qutrit state with Schmidt number at most two, together with a single positive operator M_BC, such that the post-selected state on AD is entangled; that would contradict Corollary 10.","tokens_in":13542,"feed_emoji":"🔗","tokens_out":3951,"duration_ms":39490,"temperature":0.7,"pith_summary":"The paper proves a sharp result about entanglement swapping on two-qutrit systems: if one link is 1-undistillable and the other has Schmidt number at most two, then no post-selected measurement on the intermediate systems can generate end-to-end entanglement. In map language, composing any 1-undistillable completely positive map with any 2-superpositive map is always entanglement breaking, in both orders. The authors also show the 1-undistillable cone is the largest cone with this property against the Schmidt-number-two testing cone, so the boundary is exact. This goes beyond the PPT squared conjecture by identifying a broader class -- including non-PPT yet distillable links -- that still cannot directly produce terminal entanglement in a repeater step.","feed_headline":"1-undistillable links can never swap entanglement with Schmidt-two links","feed_subtitle":"A qutrit proof shows the composition is entanglement breaking both ways, and no larger cone passes the test.","key_machinery":"The Choi-Jamiołkowski correspondence connects map composition to entanglement swapping, so that a composed map is entanglement breaking exactly when the post-selected terminal state is separable. The proof engine is Proposition 11, which characterizes 1-undistillability of a map by requiring that every rank-two local filtering of its Choi matrix be positive under partial transpose (PPT). Combined with the low-dimensional fact that every PPT operator on a 2×3 or 3×2 system is separable, this makes each term in the decomposition of the composed Choi matrix separable. The state-level extension relies on Theorem 9, which shows that when the relevant cones are closed under local filtering, provin","core_discovery":"Working with qutrit completely positive maps, the central result is an equivalence. A cone X of such maps is contained in the 1-undistillable cone UND_1 if and only if every map in X, composed with every map whose Choi matrix has Schmidt number at most two, gives an entanglement breaking map, in either order. One direction, Theorem 4, proves that composing any 1-undistillable map with any 2-superpositive map yields an entanglement breaking map for both orders of composition. The converse, Theorem 6, proves that any map outside UND_1 fails against some 2-superpositive map on at least one side. Thus UND_1 is simultaneously the maximal compatible cone and a complete test for 1-undistillability","pith_inferences":["Editorial extension: The same proof recipe could be tested in higher dimensions by replacing the testing cone with maps whose Choi matrices have Schmidt number at most k, and asking whether the compatible cone is exactly the (k-1)-undistillable class; the obstruction is that PPT operators on 2⊗d subsystems need not be separable for d>3.","Editorial extension: The maximality theorem is stated against the full testing cone SP_2; against smaller testing cones, such as PPT, the maximal compatible cone can be strictly larger, as the Werner-Holevo example shows. Mapping out this hierarchy of testing cones would complete the boundary picture.","Editorial extension: Theorem 9 is a transferable device: any pair of state cones closed under local filtering inherits the single-outcome separability guarantee for all selective measurement outcomes, so similar bridges could be reused in other repeater-based protocols."],"forward_implications":["The qutrit PPT squared conjecture becomes a special case of a broader phenomenon, since every PPT qutrit state is 1-undistillable and has Schmidt number at most two.","A 1-undistillable two-qutrit link cannot be used with any Schmidt-number-two link to produce terminal entanglement in a single selective entanglement-swapping step, in either order.","For every qutrit map outside UND_1, there is a Schmidt-number-two map that witnesses the failure of EB-composability, on at least one side of the composition.","Distillable, non-PPT entanglement in a Schmidt-number-two link does not by itself make the link directly useful as a repeater link when paired with a 1-undistillable link.","The entanglement-breaking conclusion extends from a single maximally entangled outcome to arbitrary selective measurements, because the 1-undistillable and Schmidt-number-two state cones are closed under local filtering."],"fun_headline_variants":["Qutrit swapping: 1-undistillable + Schmidt-2 always break entanglement","Entanglement swapping fails for 1-undistillable and Schmidt-2 qutrits","Composing 1-undistillable with Schmidt-2 maps gives entanglement breaking","Maximal compatible cone: exactly 1-undistillable qutrit maps","No terminal entanglement from swapping 1-undistillable with Schmidt-2 states"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof chain leans on the low-dimensional fact that every PPT operator on a 2×3 or 3×2 system is separable; if that criterion fails for the rank-two local-filtered summands used in the proof, the composition theorem no longer follows.","fun_headline_variants_meta":{"raw":{"variants":["Qutrit swapping: 1-undistillable + Schmidt-2 always break entanglement","Entanglement swapping fails for 1-undistillable and Schmidt-2 qutrits","Composing 1-undistillable with Schmidt-2 maps gives entanglement breaking","Maximal compatible cone: exactly 1-undistillable qutrit maps","No terminal entanglement from swapping 1-undistillable with Schmidt-2 states"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000963,"raw_usage":{"total_tokens":3940,"prompt_tokens":753,"completion_tokens":3187,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":497,"completion_tokens_details":{"reasoning_tokens":3077}},"tokens_in":497,"tokens_out":3187,"duration_ms":23947,"temperature":1.0,"reasoning_tokens":3077,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T21:52:59.378747+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a qutrit completely positive map with a 1-distillable Choi matrix for which composition with every Schmidt-number-two map, in both orders, is still entanglement breaking; Theorem 6 says such a map cannot exist. Alternatively, exhibit a two-qutrit state in UND_1 and a two-qutrit state with Schmidt number at most two, together with a single positive operator M_BC, such that the post-selected state on AD is entangled; that would contradict Corollary 10.","supporting_citations":[],"review_version":1}