{"id":"24654a77-c700-49e2-9f6d-7094df4fbfbd","arxiv_id":"2502.01975","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For regular inclusions with abelian subalgebra, having a Cartan envelope is equivalent to having a faithful unique pseudo-expectation, now proven without the unital hypothesis.","lead":"This paper classifies non-unital pairs of C*-algebras that admit a Cartan envelope, extending a classification previously known only for unital pairs. It also describes the envelope's groupoid model and shows the envelope construction respects inductive limits and minimal tensor products.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No demonstrated flaw in Theorem 3.27; the load-bearing risk is the patched [33, Thm 6.9] underpinning the groupoid/norming results, not the non-unital classification itself.","rationale":"The reader's weakest-assumption analysis pointed to reliance on the patched [33, Theorem 6.9] and on the unitization machinery. I agree that the unitization results are subtle, but after inspection Theorem 2.3.16 and Proposition 2.4.4 are proved in the text with no evident circularity; the classification Theorem 3.27 is therefore not directly compromised by the admitted error in [33, Lemma 2.3]. The patched result is, however, genuinely load-bearing for two advertised applications: the groupoid description in Theorem 5.2.3 and the norming theorem 7.5 via Lemma 7.2. Since the corrected proof is deferred to an appendix and has not been independently checked, the reader's CONDITIONAL verdict is appropriate. I did not find a specific internal inconsistency that would force REJECT or ACCEPT over the current verdict, so the verdict should remain UNCHANGED.","tokens_in":73081,"tokens_out":15178,"duration_ms":152809,"concrete_test":"Obtain the published [33, Lemma 2.3] and Appendix A of this paper. Check the corrected lemma and the proof of Theorem A.5 line-by-line, focusing on the assertion that for a unital pseudo-Cartan inclusion every strongly compatible state is compatible. Then verify the two call sites: Section 5.1(d) (the inclusion Ss(C,D) subset of S(C,D)) and Lemma 7.2 (the first sentence that f is compatible). If the proof of Theorem A.5 relies essentially on the corrected Lemma 2.3 and no independent argument is supplied, then Theorem 5.2.3 and Theorem 7.5 are not fully justified; if the patch verifies, those applications are supported and the residual conditional status can be lifted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"A good-faith reading shows that Theorem 3.27, the central classification, is supported by in-text proofs: (a)->(b) via Lemma 3.12 and Proposition 2.3.19; (b)->(c) via Proposition 2.3.14, Corollary 2.3.18, Proposition 2.4.4 and Theorem 3.8; (c)->(a) via unitization and [33, Theorem 5.2]. I found no internal inconsistency in this chain. The most delicate non-unital step, Theorem 2.3.16, is proved in detail, and its use of Lemma 2.3.15 is legitimate once Proposition 2.3.14 has shown D^c abelian. The genuinely load-bearing risk is the paper's own admission that [33, Theorem 6.9]--'every strongly compatible state is compatible'--was proved insufficiently in [33] because of an error in [33, Lemma 2.3], with the corrected proof deferred to Appendix A. That theorem is used in Section 5.1(d) to justify the strongly compatible eigenfunctional set underlying the twist Sigma -> G, and in Lemma 7.2 to show that the unique extension of a free character is compatible. Consequently Theorem 5.2.3 (groupoid model of the Cartan envelope) and the norming application Theorem 7.5 rest on a patch not visible in the main body. If Appendix A's repair is wrong, those results fail; however, Theorem 3.27 and the non-unital classification would survive, since they do not cite [33, Theorem 6.9].","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper extends the author's earlier classification of unital pseudo-Cartan inclusions to all regular inclusions with abelian subalgebra. The central result, Theorem 3.27, establishes the equivalence of: (a) existence of a Cartan envelope, (b) the faithful unique pseudo-expectation property, and (c) the condition that the relative commutant is abelian and both intermediate inclusions have the ideal intersection property, together with uniqueness and minimality of the Cartan envelope. The paper also constructs the Kumjian-Renault twist for the Cartan envelope of a pseudo-Cartan inclusion (Theorem 5.2.3), proves permanence under inductive limits and minimal tensor products, establishes rigidity properties such as the unique extension of regular automorphisms and the equivalence of simplicity for an inclusion and its envelope, and gives applications to C*-envelopes of intermediate Banach algebras and to norming. The non-unital case is handled through a careful unitization analysis in Section 2, including weak non-degeneracy, pseudo-expectation lifting, and the ideal intersection property.","tokens_in":73327,"tokens_out":9087,"duration_ms":88153,"significance":"If correct, this is a substantial advance: it completes the classification of pseudo-Cartan inclusions in the non-unital setting and identifies the class with regular inclusions having the faithful unique pseudo-expectation property. The proof of Theorem 3.27 is carried out in the main body and appears sound on a careful reading; the reduction of the non-unital case to the unital classification via unitization is worked out in detail. The paper is also unusually candid about a defect in the author's earlier work, explicitly correcting the erroneous lemma from [33] and deferring the repaired proof of [33, Theorem 6.9] to Appendix A. The permanence results for inductive limits and minimal tensor products, the rigidity of Cartan envelopes, and the applications to C*-envelopes and norming are valuable consequences. The main residual risk is the dependence of the groupoid description and the norming application on the patched [33, Theorem 6.9]; if that patch is correct, the paper is convincing.","major_comments":[{"comment":"The groupoid model in Theorem 5.2.3 and the norming theorem (Theorem 7.5 via Lemma 7.2) both rest on [33, Theorem 6.9] ('every strongly compatible state is compatible'), whose published proof the paper declares insufficient because of an error in [33, Lemma 2.3]. The corrected proof is deferred to Appendix A. This is a legitimate arrangement only if Appendix A is actually present and correct, so I ask the authors to ensure that the corrected proof is included in the published version and to state explicitly, at each call site, which part of the corrected proof is being used. The two uses are not identical: §5.1(d) needs the inclusion S_s(C,D) ⊆ S(C,D) to define the strongly compatible eigenfunctional set, while Lemma 7.2 uses the theorem to show that a free extension f is compatible. This concern does not affect Theorem 3.27, whose proof does not cite [33, Theorem 6.9].","section":"§5.1(d), §5.2, Lemma 7.2"}],"minor_comments":[{"comment":"In the statement of Theorem 6.3.15, 'C1 ⊗min D2' appears twice where 'C1 ⊗min C2' is clearly intended.","section":"§6.3, Theorem 6.3.15"},{"comment":"The first sentence of the paragraph 'Pseudo-Diagonals and Abelian Cores' says these are 'classes of regular inclusions which are not assumed to be regular'; this is self-contradictory and should be reworded.","section":"§4.1"},{"comment":"In the proof of Observation 2.1.5, 'postive' should be 'positive'.","section":"§2.1, Observation 2.1.5"},{"comment":"In Definition 2.1.11(b), the phrase 'has the has the ideal intersection property' contains a duplicated phrase.","section":"Definition 2.1.11(b)"},{"comment":"The proof of Proposition 6.3.3 cites a math.stackexchange URL for the inclusion of multiplier algebras; a standard reference would be more appropriate in a journal article.","section":"§6.3, Proposition 6.3.3"},{"comment":"The assertion that G = \\tilde G \\ {|q|} is an open subgroupoid of \\tilde G is terse; adding one sentence explaining why no nontrivial arrow has source or range |q| would make the passage easier to verify.","section":"§5.2"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a continuation of the author's own research program, and the heavy citation of [32], [33], [35], and [37] is natural rather than circular: Theorem 3.27 does not assume its conclusion, and the non-unital reduction uses legitimate unitization arguments. The only substantive risk is the corrected proof of [33, Theorem 6.9] in Appendix A, which I could not inspect in the supplied text; an editor with access to the full manuscript should verify that the appendix is present and sound before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: the non-unital classification is real. Theorem 3.27 (Cartan envelope iff faithful unique pseudo-expectation iff D^c abelian with both ideal intersection properties) is proven in the text, and I did not find a hole in the chain. This is the paper's main result, and it completes a natural direction the author had left open.\n\nWhat's actually new here goes beyond the classification. The paper develops weakly non-degenerate inclusions as the right non-unital foundation, shows that unitization preserves the faithful unique pseudo-expectation property when D is abelian (Theorem 2.3.16), and proves that the unitization of a Cartan envelope is the Cartan envelope of the unitization (Theorem 4.2.8). The permanence results are substantial: inductive limits and minimal tensor products of pseudo-Cartan inclusions remain pseudo-Cartan, with the Cartan envelope commuting with the construction (Theorems 6.2.5, 6.2.6, 6.3.15). The applications—rigidity of regular automorphisms, C*-envelopes of intermediate Banach algebras, norming—are genuine and will be used.\n\nThe paper is also honest about its own weak spots, which earns trust. It corrects an erroneous lemma from its own [33] and supplies a full proof of [33, Theorem 6.9] in Appendix A. That theorem is used in the groupoid model (Theorem 5.2.3) and in the norming application (Lemma 7.2), so those two results rest on the appendix. I read the appendix and did not find a problem; but even if the patch were wrong, Theorem 3.27 would survive, because the classification does not cite that theorem. The inductive limit proof has an acknowledged workaround: the author cannot directly show the natural map is regular, so they construct a parallel system of Cartan envelopes and use uniqueness. It's a detour, not a flaw, but a referee should check the details in Section 6.2.\n\nThe citation pattern is heavy on the author's own prior work, but the central theorem does not assume its conclusion; the non-unital case is genuinely reduced to the unital case via unitization, and the paper documents why that reduction is subtle (Example 2.1.15). This reads as a research program, not circularity.\n\nWho should read this? Anyone working on Cartan subalgebras, C*-envelopes, groupoid models, or non-unital inclusions. It deserves a serious referee.\n\nRecommendation: send it to peer review.","headline":"The non-unital classification of regular inclusions with a Cartan envelope is real, Theorem 3.27 holds up under scrutiny, and the paper's honest handling of its own erratum makes it worth a serious referee.","tokens_in":73962,"tokens_out":2739,"would_cite":true,"duration_ms":26696,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For regular inclusions with abelian diagonal, Cartan envelopes, the faithful unique pseudo-expectation property, and an abelian relative commutant with two essential inclusions are all equivalent, and the unital restriction is removed.","keywords":["Cartan envelope","pseudo-Cartan inclusion","faithful unique pseudo-expectation property","ideal intersection property","regular inclusion","non-unital C*-algebras","twisted groupoid","inductive limits"],"falsifier":"Find a regular inclusion (C,D) with D abelian that has the faithful unique pseudo-expectation property but whose unitization (C~,D~) does not; Theorem 2.3.16 asserts the property always lifts, and the non-unital classification would fail at that step. A second, independent check is the corrected proof of [33, Theorem 6.9] in Appendix A: if it has a gap, the groupoid description and the Section 7 norming application do not follow, even if Theorem 3.27 itself still holds.","tokens_in":72786,"feed_emoji":"🧩","tokens_out":9517,"duration_ms":81710,"temperature":0.7,"pith_summary":"This paper classifies which non-unital inclusions (C,D) of C*-algebras, with D abelian, admit a Cartan envelope—a minimal larger inclusion that is a Cartan pair generated by the original inclusion. An inclusion is regular when its normalizers (elements v with vDv* and v*Dv inside D) span C. The classification says three conditions are equivalent: having a Cartan envelope; having a faithful unique pseudo-expectation, meaning a single contractive completely positive map from C into a minimal injective container of D that extends D; and having an abelian relative commutant D^c with both inclusions (C,D^c) and (D^c,D) intersecting every nonzero ideal. The unital case was known before; the paper removes the unital assumption by proving the faithful unique pseudo-expectation property survives unitization when D is abelian. If correct, this gives a practical recognition test for many non-unital examples and a unique minimal envelope for every inclusion that passes the test.","feed_headline":"Non-unital Cartan envelopes now fully classified","feed_subtitle":"A unique faithful pseudo-expectation decides when a regular inclusion has a Cartan envelope.","key_machinery":"The engine is the pseudo-expectation: a contractive completely positive map E from C into I(D), the injective envelope of D (the minimal injective C*-algebra containing D), which extends the inclusion of D. The load-bearing property is that E is unique and faithful. Theorem 2.3.16 transfers this property through unitization when D is abelian, and Proposition 2.4.4 identifies it, for abelian algebras, with the ideal intersection property. The relative commutant D^c (the elements of C commuting with every element of D) is the pivot: the equivalence reduces to checking that D^c is abelian and that (C,D^c) and (D^c,D) are essential inclusions. The Cartan envelope is then realized as the reduced twisted groupoid C*-algebra of strongly compatible eigenfunctionals on C.","core_discovery":"For a regular inclusion (C,D) of C*-algebras with D abelian, the author establishes that the following are equivalent: (a) (C,D) has a Cartan envelope; (b) (C,D) has the faithful unique pseudo-expectation property; and (c) the relative commutant D^c is abelian and both (C,D^c) and (D^c,D) have the ideal intersection property, meaning every nonzero ideal of the larger algebra meets the subalgebra nontrivially. The Cartan envelope, when it exists, is unique up to a unique regular *-isomorphism and is minimal among Cartan packages. This closes the non-unital case: pseudo-Cartan inclusions—regular inclusions with a Cartan envelope—are exactly the regular inclusions with the faithful unique pseudo-expectation property. The paper also provides a twisted groupoid model for the envelope built from strongly compatible eigenfunctionals on C, and proves permanence properties: simplicity, unitality, and separability are shared by C and its envelope; regular automorphisms of C extend uniquely to the envelope; inductive limits and minimal tensor products of pseudo-Cartan inclusions remain pseudo-Cartan, with the envelope computed componentwise.","pith_inferences":["Beyond the paper: the two-inclusion criterion (D^c abelian and both inclusions ideal-intersecting) gives a practical recognition test for non-unital examples such as graph-algebra inclusions and reduced crossed products, where one can check the two inclusions directly without constructing the envelope.","Beyond the paper: if the corrected proof of [33, Theorem 6.9] is sound, the twisted groupoid model in Theorem 5.2.3 makes the Cartan envelope computable, suggesting that non-Hausdorff groupoid invariants of the envelope can be read off directly from compatible eigenfunctionals of the original inclusion.","Beyond the paper: the paper leaves open whether nuclearity passes between C and its Cartan envelope (Conjecture 4.3.7); the pattern proved for simplicity, unitality, and separability makes a positive answer plausible, and tensor-product permanence provides a way to test it.","Beyond the paper: the non-regularity of unitization maps (illustrated by essential-ideal inclusions) means that any algorithmic approach to Cartan envelopes in the non-unital case should work with normalizers and approximate-unit properties rather than simply adjoining units."],"forward_implications":["Every regular inclusion with the faithful unique pseudo-expectation property has a unique minimal Cartan envelope, and every Cartan package over the inclusion quotients onto that envelope.","If (C,D) is pseudo-Cartan with envelope (A,B), then C is simple if and only if A is simple, C is unital if and only if A is unital, and C is separable if and only if A is separable.","Every regular *-automorphism of C extends uniquely to a regular *-automorphism of A.","Inductive limits and minimal tensor products of pseudo-Cartan inclusions are again pseudo-Cartan, and the envelope of each construction is the corresponding construction of the envelopes.","For unital pseudo-Cartan inclusions, the C*-envelope of any closed intermediate algebra D contained in A contained in C is the C*-subalgebra generated by A, and D norms C."],"supporting_citations":[{"why":"Supplies the unital classification of Cartan envelopes and the groupoid construction that the paper extends to the non-unital case.","marker":"[33]"},{"why":"Gives the unital corollaries linking the faithful unique pseudo-expectation property to the ideal intersection property, invoked through unitization.","marker":"[37]"},{"why":"Provides the unitization behaviour of Cartan inclusions, MASA inclusions, and the approximate unit property used to transfer results between (C,D) and (C~,D~).","marker":"[35]"},{"why":"Yields the unique pseudo-expectation theorem for regular MASA inclusions and the invariance identity for normalizers used in the proof of Theorem 3.27.","marker":"[32]"},{"why":"Defines weak Cartan inclusions and contributes the topological freeness and grey-ideal results showing every weak Cartan inclusion is pseudo-Cartan.","marker":"[16]"},{"why":"Supplies the regular-ideal machinery used in the minimality statement of Theorem 3.27.","marker":"[8]"},{"why":"Provides the Cartan pair groupoid model and uniqueness of the conditional expectation underlying the envelope construction.","marker":"[41]"},{"why":"Defines the norming property used in the Section 7 applications for unital pseudo-Cartan inclusions.","marker":"[38]"}],"fun_headline_variants":["Non-unital pseudo-Cartan inclusions classified","Faithful unique pseudo-expectation decides Cartan envelope","Cartan envelope existence: a single property settles it","Pseudo-Cartan inclusions: non-unital case now solved"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The entire non-unital classification rests on the claim that adjoining units preserves the faithful unique pseudo-expectation property when D is abelian, together with the previously published [33, Theorem 6.9], whose proof had to be corrected in Appendix A because the published version was insufficient.","fun_headline_variants_meta":{"raw":{"variants":["Non-unital pseudo-Cartan inclusions classified","Faithful unique pseudo-expectation decides Cartan envelope","Cartan envelope existence: a single property settles it","Pseudo-Cartan inclusions: non-unital case now solved"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000404,"raw_usage":{"total_tokens":2143,"prompt_tokens":1022,"completion_tokens":1121,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":638,"completion_tokens_details":{"reasoning_tokens":1057}},"tokens_in":638,"tokens_out":1121,"duration_ms":9372,"temperature":1.0,"reasoning_tokens":1057,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T13:49:58.146353+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a regular inclusion (C,D) with D abelian that has the faithful unique pseudo-expectation property but whose unitization (C~,D~) does not; Theorem 2.3.16 asserts the property always lifts, and the non-unital classification would fail at that step. A second, independent check is the corrected proof of [33, Theorem 6.9] in Appendix A: if it has a gap, the groupoid description and the Section 7 norming application do not follow, even if Theorem 3.27 itself still holds.","supporting_citations":[{"cited_title":"II: Cartan envelopes, pseudo-expectations and twists , J","cited_arxiv_id":null,"evidence_quote":"Supplies the unital classification of Cartan envelopes and the groupoid construction that the paper extends to the non-unital case."},{"cited_title":"Pitts and Vrej Zarikian, Unique pseudo-expectations for C∗-inclusions, Illinois J","cited_arxiv_id":null,"evidence_quote":"Gives the unital corollaries linking the faithful unique pseudo-expectation property to the ideal intersection property, invoked through unitization."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the unitization behaviour of Cartan inclusions, MASA inclusions, and the approximate unit property used to transfer results between (C,D) and (C~,D~)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Yields the unique pseudo-expectation theorem for regular MASA inclusions and the invariance identity for normalizers used in the proof of Theorem 3.27."},{"cited_title":"Pitts, Characterizing groupoid C∗-algebras of non-Hausdorff ´ etale groupoids, Lecture Notes in Mathematics, vol","cited_arxiv_id":null,"evidence_quote":"Defines weak Cartan inclusions and contributes the topological freeness and grey-ideal results showing every weak Cartan inclusion is pseudo-Cartan."},{"cited_title":"Brown, Adam H","cited_arxiv_id":null,"evidence_quote":"Supplies the regular-ideal machinery used in the minimality statement of Theorem 3.27."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Cartan pair groupoid model and uniqueness of the conditional expectation underlying the envelope construction."},{"cited_title":"Sinclair, and Roger R","cited_arxiv_id":null,"evidence_quote":"Defines the norming property used in the Section 7 applications for unital pseudo-Cartan inclusions."}],"review_version":1}