{"id":"ed0e2123-1331-4cff-98cf-6182166489e3","arxiv_id":"2506.03845","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper characterizes generalized and pseudo n-strong Drazin invertibility via a^n - a^(2n) being quasinilpotent or lying in the Jacobson radical closure, then proves additive and weighted extensions.","lead":"This paper gives new conditions for when an element of a Banach algebra is generalized n-strong Drazin invertible and for when sums of such elements remain invertible in the same sense. It also defines weighted versions of these inverses and proves analogues for pseudo n-strong Drazin inverses.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reader's counterexample fails the hypothesis of Theorem 3.8; the real gap is the unjustified step that the two summands in the decomposition lie in p√J.","rationale":"The reader and I converge on Theorem 3.8 as the key weak point, but for different reasons. The reader's proposed counterexample a=e12, b=e21 does not satisfy the hypothesis ab^2=bab, and the identity in question actually does hold under that hypothesis; that part of the reader's analysis is incorrect. However, the written proof of Theorem 3.8 is still incomplete: it moves from equation (1) to the conclusion without proving that the two summands belong to p√J, a step that is not a consequence of the general facts stated in Section 3. Since p√J is not closed under one-sided multiplication, this is a real gap rather than a stylistic shortcut. The central characterizations (Theorems 2.4, 3.5 and their corollaries) appear sound and are independent support for the paper's main novelty. The additive and product claims, especially Theorem 3.8 and the related Theorem 3.17, require additional argument before they can be accepted as proved. Therefore the conditional verdict is appropriate, though the stated justification should be revised.","tokens_in":14568,"tokens_out":36674,"duration_ms":405025,"concrete_test":"For n=2, search over 3×3 complex matrices (or a finite candidate set) for pairs (a,b) satisfying ab^2=bab, with a-a^3 and b-b^3 nilpotent. For each such pair, compute y=a^3(b-b^3) and check whether y is nilpotent. If any pair gives non-nilpotent y, Theorem 3.8 is false; if none does, the theorem may still be true but the proof needs a new lemma justifying y∈p√J. As an additional check, verify whether ab-(ab)^3 is nilpotent for all found pairs; a counterexample there would directly disprove Theorem 3.8.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The specific check proposed in the reader's verdict does not land: with a=e12 and b=e21 in M2, one gets ab^2=0 but bab=e21, so ab^2=bab fails. Moreover, the identity (ab)-(ab)^{n+1}=a^{n+1}(b-b^{n+1})+(a-a^{n+1})b is valid under ab^2=bab, because this relation implies (ab)^m=a^m b^m for all m by the induction b^m a b = a b^{m+1}. The genuine weakness is elsewhere: after writing ab-(ab)^{n+1}=X+Y with X=a^{n+1}(b-b^{n+1}) and Y=(a-a^{n+1})b, the proof only cites Theorem 3.5/Corollary 3.6. From a,b in ApnsD we get a-a^{n+1}, b-b^{n+1} in p√J, but nothing in the written proof shows X,Y themselves lie in p√J. This is not automatic: p√J is not closed under left or right multiplication by arbitrary elements (in M2 over C, e12∈p√J but e21e12=e22∉p√J). No argument is given that the specific multipliers a^{n+1} and b preserve p√J under the hypothesis ab^2=bab, nor that X and Y satisfy the product/commutation hypotheses needed to apply Lemma 3.2 or Lemma 3.3. Thus Theorem 3.8 is not established as written, and the additive equivalence Theorem 3.17, which is modeled on the analogous gns argument, inherits this incompleteness.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces and studies generalized n-strong Drazin (gns) and pseudo n-strong Drazin (pns) invertibility in Banach algebras. The main structural results are the characterizations of Theorem 2.4 (a is gns-invertible if and only if a^n - a^{2n} is quasinilpotent) and Theorem 3.5 (a is pns-invertible if and only if a^n - a^{2n} lies in the radical closure psqrt(J(A))), together with the corollaries a - a^{n+1} quasinilpotent or in psqrt(J(A)). These results are then applied to prove product and additive theorems, including the equivalence statements in Theorems 2.20 and 3.17, and weighted analogues in Section 4.","tokens_in":14930,"tokens_out":23217,"duration_ms":226665,"significance":"The main characterizations are clean, potentially useful, and appear to be correct; they extend earlier results of Chen-Sheibani and Mosic in a natural direction. The manuscript contains no fitted parameters or circular definitions, and the proofs of the characterizations rely on standard spectral mapping and radical arguments. However, several load-bearing additive and product theorems are not established as written because of missing justifications and a sign error. If the gaps are repaired, the paper would make a solid contribution to the generalized Drazin inverse literature.","major_comments":[{"comment":"The proof of Theorem 3.8 is the single sentence \"It follows from equation 1 and Theorem 3.5.\" This is not sufficient. Corollary 3.6 gives a - a^{n+1} and b - b^{n+1} in psqrt(J(A)), but equation (1) expresses (ab) - (ab)^{n+1} as a^{n+1}(b - b^{n+1}) + (a - a^{n+1})b. The proof never shows that these two summands lie in psqrt(J(A)) or satisfy the hypotheses of Lemma 3.2 or Lemma 3.3. psqrt(J(A)) is not closed under multiplication by arbitrary elements, and no argument is given that the particular multipliers a^{n+1} and b preserve it under the hypothesis ab^2 = bab. The concrete example a = e12, b = e21 does not satisfy the hypothesis ab^2 = bab, so it is not a counterexample to the theorem; nevertheless the proof gap is real. Consequently Theorem 3.8, and any result that depends on it, is not established as written.","section":"Section 3, Theorem 3.8"},{"comment":"The expansion in the proof of Theorem 2.9 has the wrong sign. The displayed identity reads (a+b) - (a+b)^{n+1} = (a - a^{n+1}) + (b - b^{n+1}) + sum_{i=1}^n b^i a^{n+1-i}, but under ab = 0 the mixed term should be subtracted, not added. The subsequent appeal to Lemma 2.2 is therefore not justified for the expression as written. The same sign error appears in the proof of Theorem 3.9. Since Theorem 2.9 is used in later arguments, this error propagates and must be corrected.","section":"Theorems 2.9 and 3.9"},{"comment":"The proof of Theorem 2.20 contains unsupported steps that need to be supplied. In particular, the claim that b4 is in Ansd_2 is inferred from a a^nsd b being in Ansd, although that product only sees the (1,1) block of b; an additional spectral or block-matrix argument is needed. Later, the step \"by Lemma 2.7\" from p1 + (b41)^{-1}a21 in Ansd to a21 + b41 in Ansd requires both factors of the product to be gns-invertible, but invertibility alone does not imply gns-invertibility. The statement of Theorem 3.17 says it follows \"as Theorem 2.20\" and therefore inherits these unresolved points. These are gaps in the proof of the paper's main additive equivalence results, not merely presentation issues.","section":"Theorem 2.20 and Theorem 3.17"}],"minor_comments":[{"comment":"There are evident subscript typos: in Lemma 3.11(2) the conclusion should refer to ApnsD_1 and ApnsD_2 in the correct order, and in Lemma 3.12 the statement that \"b1 is in ApnsD_2\" should say ApnsD_1, since b1 is an element of the corner A1.","section":"Lemma 3.11(2) and Lemma 3.12"},{"comment":"The displayed definition of the g-Drazin inverse uses the condition b = ab^2. This is nonstandard and may confuse readers; it is equivalent to the usual b = bab only when combined with ab = ba, so the equivalence should be stated explicitly.","section":"Introduction, definition of g-Drazin inverse"},{"comment":"The notation \"gns\" and \"pns\" is typeset inconsistently (e.g., \"g ns-invertibility\" in the abstract and \"gns-Drazin invertible\" in the body). The paper would benefit from consistent formatting.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper's central characterizations are sound and likely publishable once the additive and product theorems are fully proved. The main work for the revision is to provide complete proofs for Theorems 3.8, 3.17, and the questionable steps in Theorem 2.20. The sign errors in Theorems 2.9 and 3.9 appear easily fixable but must be addressed because the current proofs are incorrect as written."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Rounak and Falguni's paper is a serious contribution to the generalized-inverse literature. The central results—Theorem 2.4 and Theorem 3.5—are clean characterizations: a is gns-Drazin invertible iff a^n − a^{2n} is quasinilpotent, and pns-invertible iff the same expression lies in √J(A). These are new, they unify earlier special cases, and the proofs check out. That alone is worth something.\n\nThe additive sections are where I have problems. The sign errors in Theorem 2.9 and Theorem 3.9 are real but minor: the expansion of (a+b)−(a+b)^{n+1} should have a minus sign on the mixed term. With ab=0 the corrected term is still quasinilpotent/radical, so the theorem survives a simple fix.\n\nTheorem 3.8 is more serious. The reader's report claims equation (1) fails under ab^2=bab, with an e12/e21 example. That example does not satisfy ab^2=bab, and the identity is actually correct under that hypothesis—it follows from (ab)^m = a^m b^m. So the stress-test note is right to dismiss that counterexample. But the stress-test also identifies the real gap: after writing ab−(ab)^{n+1}=X+Y, the proof gives no reason why X=a^{n+1}(b−b^{n+1}) and Y=(a−a^{n+1})b lie in √J(A). The radical √J(A) is not closed under multiplication by arbitrary elements, and the hypothesis ab^2=bab is not used to show those specific products remain in √J. The one-line proof \"It follows from equation 1 and Theorem 3.5\" is not enough. As a result, Theorem 3.8 is not established as written, and Theorem 3.17 inherits the gap because it uses the same decomposition pattern.\n\nNone of this is fatal to the main characterizations. The paper is honest, the bibliography is standard, and the weighted section in Section 4 is a reasonable extension. The right outcome is a revise-and-resubmit, not a rejection. I would send it to a referee who knows the area and will insist on a full proof of Theorem 3.8 and the two sign fixes. It belongs in a specialized journal. I would bring it to our reading group, if only to talk about what a correct proof of Theorem 3.8 needs.","headline":"Useful characterizations with a repairable additive gap: the main theorem pair is solid, but Theorem 3.8 needs a real proof fix before the additive claims are accepted.","tokens_in":15440,"tokens_out":4217,"would_cite":true,"duration_ms":39951,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["15A09","16N20"],"pacs":[],"model":"deepseek-v4-flash","headline":"In a Banach algebra, an element is generalized n-strong Drazin invertible exactly when $a^n-a^{2n}$ is quasinilpotent, and pseudo n-strong Drazin invertible exactly when $a-a^{n+1}$ lies in the radical closure $p\\sqrt{J(A)}$.","keywords":["generalized n-strong Drazin inverse","pseudo n-strong Drazin inverse","additive property","Banach algebra","Jacobson radical","quasinilpotent","weighted Drazin inverse"],"falsifier":"Compute both sides of the identity $(ab)-(ab)^{n+1}=a^{n+1}(b-b^{n+1})+(a-a^{n+1})b$ in a Banach algebra under the hypothesis $ab^2=bab$. In $M_2(\\mathbb{C})$ with $a=e_{12}$, $b=e_{21}$, and $n=2$, the left side is $0$ and the right side is $e_{11}$, so the identity is not a consequence of the stated hypothesis. To falsify Theorem 3.8 as a statement, one would need a pair of pns-invertible elements satisfying $ab^2=bab$ for which $ab$ is not pns-invertible.","tokens_in":14379,"feed_emoji":"🧮","tokens_out":6246,"duration_ms":57010,"temperature":0.7,"pith_summary":"This paper establishes a single spectral test for two flavours of n-strong Drazin invertibility in a Banach algebra: an element $a$ is generalized n-strong Drazin invertible precisely when $a^n-a^{2n}$ is quasinilpotent, equivalently when $a-a^{n+1}$ is quasinilpotent, and pseudo n-strong Drazin invertible precisely when $a^n-a^{2n}$, equivalently $a-a^{n+1}$, lies in the radical closure $p\\sqrt{J(A)}$. With these characterizations the authors prove additive results: conditions under which $a+b$ is gns- or pns-invertible, including an equivalence theorem tying $a+b$, $a\\cdot a^{\\mathrm{nsd}}\\cdot(a+b)$, $(a+b)\\cdot b^{\\mathrm{nsd}}$, and the two-sided product. The paper then defines weighted gns and pns inverses and shows they reduce to ordinary ones by multiplying on either side by the weight.","feed_headline":"A single spectral test decides n-strong Drazin invertibility","feed_subtitle":"New characterizations and additive rules for generalized and pseudo n-strong Drazin inverses in Banach algebras.","key_machinery":"The main machinery is the polynomial $p(x)=x-x^{n+1}$, used with the spectral mapping theorem to force the spectrum of a gns-invertible element into the roots of unity together with zero, making the spectral point $0$ isolated and the element g-Drazin invertible. The additive proofs rely on block-matrix decompositions relative to spectral idempotents, on lemmas saying that a sum of quasinilpotent (or radical-closure) elements remains quasinilpotent (or radical-closure) under weak commutation conditions, and on the identity $(ab)-(ab)^{n+1}=a^{n+1}(b-b^{n+1})+(a-a^{n+1})b$, which was derived in Lemma 2.7 under the hypothesis $a^2b=aba$.","core_discovery":"The central claim is that generalized n-strong Drazin invertibility and pseudo n-strong Drazin invertibility are each characterized by one condition on powers of the element, with no separate Drazin-invertibility hypothesis needed. For sums, Theorem 2.20 states a list of equivalences under the mixed commutation conditions $a^2b=aba$ and $ab^2=bab$, and Theorem 3.17 is the pseudo analogue. The weighted version asserts that $a$ is wgns-invertible if and only if $wa$ and $aw$ are gns-invertible, with the explicit formula $a^{\\mathrm{nsd},w}=((aw)^{\\mathrm{nsd}})^2 a$, and similarly for the pseudo weighted inverse.","pith_inferences":["The pseudo product theorem (Theorem 3.8) is not established as written: its proof invokes the identity from Lemma 2.7, which was proved under $a^2b=aba$, while the theorem assumes only $ab^2=bab$.","A concrete check shows the identity is not a consequence of the weaker hypothesis: in $M_2(\\mathbb{C})$ with $a=e_{12}$, $b=e_{21}$, and $n=2$, the left side $(ab)-(ab)^{n+1}$ is $0$ while the right side $a^{n+1}(b-b^{n+1})+(a-a^{n+1})b$ equals $e_{11}$.","Repairing the pseudo product theorem likely requires adding $a^2b=aba$ to its hypotheses or finding a different identity that holds under $ab^2=bab$ alone.","The weighted characterization suggests that in operator-algebra settings, weighted n-strong Drazin invertibility can be tested by ordinary invertibility after one-sided multiplication by the weight, which may simplify computations."],"forward_implications":["Checking gns or pns invertibility reduces to a single quasinilpotence or radical-closure test on powers of the element, instead of verifying the full Drazin equations.","The additive equivalence theorems give four interchangeable formulations for $a+b$, so establishing any one of them certifies the other three.","Weighted gns and pns inverses are explicitly constructible from ordinary ones via $((aw)^{\\mathrm{nsd}})^2 a$ and $((aw)^{\\mathrm{pnsD}})^2 a$.","The results extend earlier sum theorems for generalized Drazin and generalized strong Drazin inverses to the n-strong setting, including the pseudo case.","Examples show that $a+b$ can be n-strong Drazin invertible without $ab$ being quasinilpotent, so the additive characterizations are genuinely broader than a simple product condition."],"supporting_citations":[{"why":"Defines generalized n-strong and pseudo n-strong Drazin inverses in rings, the objects this paper characterizes in Banach algebras.","marker":"[9]"},{"why":"Proved the gs-Drazin invertibility characterization that Corollary 2.5 refines and generalizes.","marker":"[11]"},{"why":"Proved the g-Drazin sum equivalences that Theorem 2.20 and Theorem 3.17 extend to the n-strong setting.","marker":"[3]"},{"why":"Supplies lemmas on Jacobson radical elements and pseudo Drazin inverse sum results used throughout Section 3.","marker":"[18]"},{"why":"Provides additive criteria for g-Drazin invertibility that Theorems 2.14 and 2.17 adapt to gns inverses.","marker":"[15]"},{"why":"Gives the quasinilpotence lemmas with commutativity up to a factor that support the main gns arguments.","marker":"[5]"}],"fun_headline_variants":["Single power condition decides n-strong Drazin invertibility","Additive formulas for generalized and pseudo n-strong Drazin inverses","Weighted n-strong Drazin inverses characterized via explicit formula","One test for n-strong invertibility, no Drazin hypothesis required"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The product proof for pseudo n-strong Drazin inverses relies on an identity previously proved under the condition $a^2b=aba$, while the theorem's stated hypothesis is only $ab^2=bab$; that identity can fail under the weaker condition.","fun_headline_variants_meta":{"raw":{"variants":["Single power condition decides n-strong Drazin invertibility","Additive formulas for generalized and pseudo n-strong Drazin inverses","Weighted n-strong Drazin inverses characterized via explicit formula","One test for n-strong invertibility, no Drazin hypothesis required"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00073,"raw_usage":{"total_tokens":3232,"prompt_tokens":872,"completion_tokens":2360,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":488,"completion_tokens_details":{"reasoning_tokens":2281}},"tokens_in":488,"tokens_out":2360,"duration_ms":20332,"temperature":1.0,"reasoning_tokens":2281,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:08:39.367058+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute both sides of the identity $(ab)-(ab)^{n+1}=a^{n+1}(b-b^{n+1})+(a-a^{n+1})b$ in a Banach algebra under the hypothesis $ab^2=bab$. In $M_2(\\mathbb{C})$ with $a=e_{12}$, $b=e_{21}$, and $n=2$, the left side is $0$ and the right side is $e_{11}$, so the identity is not a consequence of the stated hypothesis. To falsify Theorem 3.8 as a statement, one would need a pair of pns-invertible elements satisfying $ab^2=bab$ for which $ab$ is not pns-invertible.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines generalized n-strong and pseudo n-strong Drazin inverses in rings, the objects this paper characterizes in Banach algebras."},{"cited_title":"On gs-drazin inverses in a ring, Journal of Algebra and Its Applications,","cited_arxiv_id":null,"evidence_quote":"Proved the gs-Drazin invertibility characterization that Corollary 2.5 refines and generalizes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proved the g-Drazin sum equivalences that Theorem 2.20 and Theorem 3.17 extend to the n-strong setting."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies lemmas on Jacobson radical elements and pseudo Drazin inverse sum results used throughout Section 3."},{"cited_title":"134 (2004) 1085–1097","cited_arxiv_id":null,"evidence_quote":"Provides additive criteria for g-Drazin invertibility that Theorems 2.14 and 2.17 adapt to gns inverses."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the quasinilpotence lemmas with commutativity up to a factor that support the main gns arguments."}],"review_version":1}