{"id":"adbc9196-b313-4b02-907d-029c6b81a9b1","arxiv_id":"2506.01647","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For odd-dimensional Callias-type operators whose potential need not be invertible at infinity, the author derives a functional equation linking two higher-order spectral shift functions and proves a regularized index formula.","lead":"This paper develops a framework of higher-order spectral shift functions for non-Fredholm Callias-type Dirac operators in odd dimensions, and derives a functional equation that generalizes Pushnitski's one-dimensional formula. It defines a partial Witten index and computes explicit spectral shift functions for massless (d+1)-Dirac-Schrödinger operators.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Main functional equation rests on an unproved, self-cited trace formula (Theorem 3.4, from [17]); if that formula's hypotheses or proof fail, the spectral shift construction collapses.","rationale":"The reader's weakest assumption identifies exactly the same load-bearing concern: the central trace formula is imported from an unpublished, self-cited preprint without proof. I agree that this is the single most important unverified premise. The paper's own contribution -- the construction of higher-order spectral shift functions and the derivation of the functional equation -- is conditional on Theorem 3.4. If Theorem 3.4 has an error or requires hypotheses not implied by Hypothesis 3.7, then the functional equation (3.63), the partial Witten index definition, and Theorem 4.4 all fail. The Section 5 example provides one nontrivial class satisfying Hypothesis 3.7, but it does not independently prove the trace formula; it relies on [19] for that. A second, more internal concern is the application of Hypothesis 2.16 in Proposition 3.6, where the condition m >= alpha*n with m=N may conflict with the parameter choice in the massless example (N=1, alpha>1, d>=3); however, this is harder to assess without the full details of the operator norms, and the imported theorem remains the more fundamental external dependency. The proposed concrete test -- either obtaining and checking the proof of [17, Thm 6.5] or verifying the trace formula in a simple solvable case -- would settle whether the concern lands. Since the reader already returned CONDITIONAL and my read does not alter that, the verdict should remain unchanged.","tokens_in":37657,"tokens_out":14630,"duration_ms":140489,"concrete_test":"Obtain the full proof of [17, Theorem 6.5] (or its published version) and check that its hypotheses are no stronger than Hypothesis 3.2/3.7, specifically: (i) the trace-class justification of tr_{C^r}(e^{-tD*D} - e^{-tDD*}) on L^2(R^d,H) under (3.6)-(3.8); (ii) the finiteness and phi-independence of the x-integral in (3.9). If [17] requires extra assumptions, exhibit a B satisfying Hypothesis 3.7 but violating one of them; if such B exists, the domain of the main theorem shrinks. Alternatively, verify Theorem 3.4 in the exactly solvable case d=3, H=C, A0=0, B=m*1 + chi_R(x) with m>0 and chi_R a compactly supported bump, by computing both sides of (3.9) explicitly for small t.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central results (Theorem 3.14 functional equation and Theorem 4.4 index formula) are derived by feeding the principal trace formula of [17, Thm 6.5], restated as Theorem 3.4, into the abstract spectral shift machinery. This formula is not proved here; it is imported from an unpublished, self-cited preprint. The formula is delicate: after the partial trace tr_{C^r}, the semigroup difference e^{-tD*D} - e^{-tDD*} is asserted to be trace class on L^2(R^d,H), and the x-integral in (3.9) is asserted to be finite and independent of the cut-off phi. Any unstated condition in [17] -- on decay of nabla A, on the domain of A0, or on the commutator structure needed for the Kato-Rellich argument -- would invalidate the input to Propositions 3.6 and 3.13 and hence the functional equation (3.63) and Theorem 4.4. The Section 5 example verifies Hypothesis 3.7 for one class of potentials, but it does not independently verify Theorem 3.4 for that class, since the trace formula there is again cited from [19]. Thus the central claim is only as secure as an unpublished preprint.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Dirac-Schrödinger operators on odd-dimensional Euclidean space with operator-valued potentials, in the spirit of Callias but without assuming invertibility of the potential at infinity. Under a technical Hypothesis 3.7, the author constructs higher-order spectral shift functions ξ and η, derives a functional equation (3.63) generalizing Pushnitski's one-dimensional formula, defines a 'partial Witten index' as -ξ^{(d-1)}(0+), and proves an index formula (4.14) under a Lebesgue point condition on η. The final section computes ξ and η explicitly for the (d+1)-massless Dirac-Schrödinger operator and recovers the Witten index formula of [19].","tokens_in":37923,"tokens_out":26335,"duration_ms":261886,"significance":"If the technical hypotheses are satisfied, this would be the first multi-dimensional non-Fredholm extension of Callias index theory via higher-order spectral shift functions. The paper is clearly organized and the abstract construction of spectral shift measures through multiple operator integrals is a useful framework. The explicit formulas for the massless example, expressed through Bessel kernels and an index density, are concrete and checkable. However, the central results are conditional on an unproved imported trace formula and on parameter conditions that are internally inconsistent in their current form, so the significance is not yet fully established.","major_comments":[{"comment":"The principal trace formula (3.9) is quoted from the author's unpublished preprint [17, Theorem 6.5] and is not proved or verified in the present paper. Since Propositions 3.6 and 3.13, and hence Theorem 3.14 and Theorem 4.4, are all derived by feeding this formula into the spectral-shift machinery, any error or unstated condition in [17] would invalidate the central claims. The paper should either include a proof of (3.9), or state and verify all hypotheses of [17, Theorem 6.5] in the present setting, or explicitly present the results as conditional on that preprint.","section":"Section 3, Theorem 3.4"},{"comment":"Hypothesis 2.16 assumes m ≥ α n, but Proposition 2.11 and Corollary 2.10, on which Theorem 2.20 relies, require m ≥ α(n+1). Theorem 2.20 invokes Proposition 2.11 while only assuming the weaker inequality. This gap propagates to Propositions 3.6 and 3.13, where m=N, n=d-1 or d, and α>1; the required inequality then forces N to be at least of order d, whereas the example in Section 5 takes N=1 for all odd d≥3. Please reconcile the abstract hypotheses and adjust the example or the parameter choices accordingly.","section":"Section 2, Hypothesis 2.16 and Theorem 2.20"},{"comment":"The proof of Proposition 4.1 asserts that ∫_0^∞ (-t)^d e^{-tλ}ξ(λ)dλ equals ∫_0^∞ (-t)e^{-tλ}ξ^{(d-1)}(λ)dλ, but for d odd this integration by parts produces boundary terms involving ξ^{(j)}(0+) for 0≤j≤d-2. These boundary terms are not assumed to vanish and do not follow from the stated hypotheses. For example, with d=3 and η(μ)=μ^{-1/2} near 0, the functional equation gives ξ(λ) proportional to λ, so ξ'(0+)≠0 while ξ''=0; the two integrals then differ. Thus the semigroup limit formula (4.1) is not established. Please add and verify vanishing conditions on the lower derivatives of ξ at 0, or modify the statement.","section":"Section 4, Proposition 4.1"},{"comment":"The verification that the massless example satisfies Hypothesis 3.7 is delegated to [19, Lemma 2.5] with the comment 'almost ad verbatim', and the trace formula used in the example is again cited from [19, Theorem 3.5]. Since Hypothesis 3.7 is a long list of technical conditions and this example is the only evidence of non-vacuity, a direct verification or at least a detailed lemma with all required estimates should be provided. In particular, the parameter choice N=1 needs to be reconciled with the m ≥ α(n+1) condition from the abstract framework.","section":"Section 5, Theorem 5.3"}],"minor_comments":[{"comment":"In the displayed formula for the integral over the simplex, the integration variable is written as d u but should be d s.","section":"Section 5, equation before (5.26)"},{"comment":"The phrase 'for α > 1 small enough such that N = 1' is confusing because the same symbol α is used in Hypothesis 3.2 and in Definition 2.6; in light of the parameter mismatch discussed above, this choice needs to be justified or revised.","section":"Section 5, Theorem 5.3"},{"comment":"The reduction 'without loss of generality B ≡ 0 near 0' is plausible, but it should be spelled out explicitly that the spectral shift functions ξ and η are independent of the cut-off φ and that the operators D_B and D_{B_φ} give the same trace formula; otherwise the reader cannot follow the reduction.","section":"Section 3, after Lemma 3.8"},{"comment":"The strong measurability argument in Lemma 2.14 is very terse; the convergence statements involving K_N and P_k would benefit from a few more details, especially where Lemma 2.19 is used.","section":"Section 2, Lemma 2.14"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is heavily self-referential: the central input [17] is an unpublished preprint by the same author, and the example is verified by reference to the author's published paper [19]. For a journal publication, I would advise the editor to require either a full proof of Theorem 3.4 or an explicit statement that the results are conditional on [17]. The parameter inconsistency in Section 2 appears fixable but requires substantial reworking of the example and of the abstract hypotheses."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know. First, the paper does what it says: it constructs higher-order spectral shift functions for Callias-type operators in odd dimensions without the Fredholm assumption, proves a functional equation that reduces to Pushnitski's in d=1, and in the massless example computes both ξ and η explicitly. The Section 2 framework for spectral shift measures with measurable dependence on parameters is carefully built and looks correct. The Laplace inversion leading to (3.63) is coherent, and the partial Witten index formula (4.14) follows with clean analytic arguments.\n\nThe soft spot is exactly where the stress-test puts it: the principal trace formula, Theorem 3.4, is imported from the author's unpublished arXiv preprint [17] and is not proved here. Everything downstream—the definition of η in Proposition 3.6, the construction of Ξ in Proposition 3.13, and hence the functional equation and index theorem—rests on that formula. The paper is transparent about this and the cited preprint is public, so it is not a hidden dependency. But it is load-bearing: if [17] has an unstated condition or a gap, the new results collapse. Also, Hypothesis 3.7 is only shown to hold in the massless example, and that verification itself relies on [19]. So the general non-vacuity of the main hypothesis is open.\n\nI do not see a flaw in the internal logic. I checked the sign algebra around (3.64)–(3.66) and the cancellation of 1/d factors; it works. The Fredholm consistency check in Lemma 4.3 is a nice touch and the d=1 comparison is honest and useful.\n\nVerdict: worth a serious referee, conditionally. The referee should be asked to verify [17] (or the author should incorporate its proof as an appendix) and to expand the class of examples satisfying Hypothesis 3.7 beyond the massless case. This is a real contribution to non-Fredholm index theory, not a desk reject.","headline":"Genuine extension of Pushnitski's functional equation to odd-dimensional non-Fredholm Callias operators, with explicit spectral shift functions in the massless example, but the load-bearing trace formula is imported from an unpublished preprint.","tokens_in":38411,"tokens_out":7182,"would_cite":true,"duration_ms":66034,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["47A55","47B10","47A60","58J20"],"pacs":[],"model":"deepseek-v4-flash","headline":"Higher-order spectral shift functions extend Callias index theory to non-Fredholm Dirac operators.","keywords":["Callias index theorem","higher-order spectral shift functions","non-Fredholm operators","Witten index","Dirac-Schrödinger operators","multiple operator integrals","functional equation","massless Dirac operators"],"falsifier":"Take a $d=3$ potential satisfying the weaker Hypothesis 3.2 but not the stronger Hypothesis 3.7, compute both sides of the principal trace formula (Theorem 3.4) numerically for small $t>0$, and check whether the identity holds; if it fails, the functional equation and index theorem have no foundation.","tokens_in":37460,"feed_emoji":"📐","tokens_out":7969,"duration_ms":83139,"temperature":0.7,"pith_summary":"Callias's index theorem computes the Fredholm index of a Dirac-Schrödinger operator from the asymptotic winding of its potential. This paper removes the key Fredholm assumption—that the potential is invertible outside a compact set—and still obtains an index-type invariant in odd dimensions. The engine is a pair of higher-order spectral shift functions, $\\eta$ and $\\xi$, built from multiple operator integrals; they satisfy a functional equation that generalizes Pushnitski's one-dimensional formula. Under a Lebesgue point condition on $\\eta$, the operator has a well-defined partial Witten index equal to $(4\\pi)^{-(d-1)/2}L$, and when the operator is Fredholm this index agrees with the classical Fredholm index. The author claims this is the first multi-dimensional non-Fredholm extension of the Callias index theorem, and works out the $(d+1)$-massless Dirac-Schrödinger example explicitly.","feed_headline":"Callias index theory survives without Fredholmness","feed_subtitle":"Higher-order spectral shift functions give a partial Witten index that matches the Callias index when Fredholm","key_machinery":"The load-bearing machinery is the pair of higher-order spectral shift functions $\\eta_{d,A_0,B}$ and $\\xi_{d,A_0,B}$, constructed through multiple operator integrals with divided-difference symbols and the Potapov–Skripka–Sukochev trace-norm estimate (Theorem 2.7). The identity that carries the argument is the functional equation (3.63), which expresses $\\xi$ as a fractional integral of $\\eta$; it converts the cited principal trace formula into the index theorem. The partial Witten index is then read off as the right Lebesgue value $-\\xi^{(d-1)}(0+)$.","core_discovery":"The paper's central claim is that the regularized content of a Euclidean Callias-type operator survives when the potential is not invertible at infinity: the semigroup difference $\\operatorname{Tr}\\operatorname{tr}_{\\mathbb{C}^r}(e^{-tD^*D}-e^{-tDD^*})$ admits a density $\\xi_{d,A_0,B}$, the potential-side trace admits a density $\\eta_{d,A_0,B}$, and these densities are linked by the exact fractional-integral relation (3.63). From this relation, under the Lebesgue point condition (4.13), the partial Witten index exists and equals $(4\\pi)^{-(d-1)/2}L$; if $D_B$ is Fredholm, this equals the ordinary index. The discovery is that a non-Fredholm Callias-type operator still has a well-defined, computable index invariant, expressed through higher-order spectral shift functions, and the paper computes these functions for the massless example via an index density and Bessel-function kernels.","pith_inferences":["A natural testable extension is whether the partial Witten index is independent of the chosen regularization (semigroup versus resolvent) in higher odd dimensions; the paper only defines it through the semigroup limit.","The explicit Bessel-kernel formulas suggest a local, density-type picture: the higher-order spectral shift functions may be computable from pointwise index densities for a wider class of non-Fredholm potentials than the massless model, which would amount to a non-Fredholm local index theorem.","Because the index formula (5.16) depends only on the unitary evolution $U^V$ and its exterior derivative, one could test stability of the partial Witten index under compactly supported perturbations of $V$; the paper does not address this explicitly."],"forward_implications":["For any odd dimension $d$, the functional equation (3.63) gives the first known dimensional generalization of Pushnitski's spectral-shift formula, and in $d=1$ it reduces to it after identifying $\\eta$ with the symmetrized Krein spectral shift.","Whenever the potential-side spectral shift $\\eta$ satisfies the Lebesgue point condition, the partial Witten index exists and is given by the explicit constant $(4\\pi)^{-(d-1)/2}L$, providing a computable substitute for the Callias index without Fredholmness.","In the Fredholm case the partial Witten index coincides with the ordinary Fredholm index, so the new invariant is a genuine extension rather than a replacement.","For $(d+1)$-massless Dirac-Schrödinger operators, both spectral shift functions are given explicitly as integral transforms of an index density, and the partial Witten index equals the exterior-product formula (5.16)."],"supporting_citations":[{"why":"Supplies the principal trace formula (Theorem 3.4) expressing the semigroup difference as an integral over the potential; the paper's later results are derived from it.","marker":"[17]"},{"why":"Establishes existence of higher-order spectral shift functions via the multiple-operator-integral estimate used as the technical backbone of the construction.","marker":"[31]"},{"why":"Provides the one-dimensional Pushnitski functional equation that (3.63) generalizes to odd dimensions.","marker":"[32]"},{"why":"Gives the one-dimensional Witten-index result via Lebesgue points of spectral shift functions that Theorem 4.4 extends.","marker":"[13]"},{"why":"Introduces the Witten index which the paper adapts to a partial Witten index through the partial trace.","marker":"[22]"},{"why":"Prior work on the Witten index of massless $(d+1)$-Dirac-Schrödinger operators whose example and index formula are extended to full spectral shift functions.","marker":"[19]"},{"why":"Original Callias index theorem that the paper's non-Fredholm index formula generalizes.","marker":"[11]"}],"fun_headline_variants":["Non-Fredholm Callias operators still get an index","Higher-order spectral shift salvages Callias index","Callias index without Fredholmness via spectral shift","Regularized index for non-Fredholm Callias operators","Spectral shift functions extend Callias index to non-Fredholm"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the principal trace formula taken from the author's earlier preprint and reproduced as Theorem 3.4—the semigroup difference equals an explicit integral over the potential—together with the technical Hypothesis 3.7, both assumed without proof beyond the massless example.","fun_headline_variants_meta":{"raw":{"variants":["Non-Fredholm Callias operators still get an index","Higher-order spectral shift salvages Callias index","Callias index without Fredholmness via spectral shift","Regularized index for non-Fredholm Callias operators","Spectral shift functions extend Callias index to non-Fredholm"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000603,"raw_usage":{"total_tokens":2868,"prompt_tokens":1053,"completion_tokens":1815,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":669,"completion_tokens_details":{"reasoning_tokens":1735}},"tokens_in":669,"tokens_out":1815,"duration_ms":13103,"temperature":1.0,"reasoning_tokens":1735,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T11:37:44.688609+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a $d=3$ potential satisfying the weaker Hypothesis 3.2 but not the stronger Hypothesis 3.7, compute both sides of the principal trace formula (Theorem 3.4) numerically for small $t>0$, and check whether the identity holds; if it fails, the functional equation and index theorem have no foundation.","supporting_citations":[{"cited_title":"Trace and Index of Dirac-Schr\\\"odinger Operators on Open Space with Operator Potentials","cited_arxiv_id":"2311.02593","evidence_quote":"Supplies the principal trace formula (Theorem 3.4) expressing the semigroup difference as an integral over the potential; the paper's later results are derived from it."},{"cited_title":"Potapov, A","cited_arxiv_id":null,"evidence_quote":"Establishes existence of higher-order spectral shift functions via the multiple-operator-integral estimate used as the technical backbone of the construction."},{"cited_title":"Pushnitski, The spectral flow, the Fredholm index, and the spectral shift function, Spectral Theory of Differential Operators: M","cited_arxiv_id":null,"evidence_quote":"Provides the one-dimensional Pushnitski functional equation that (3.63) generalizes to odd dimensions."},{"cited_title":"Carey, F","cited_arxiv_id":null,"evidence_quote":"Gives the one-dimensional Witten-index result via Lebesgue points of spectral shift functions that Theorem 4.4 extends."},{"cited_title":"Gesztesy and B","cited_arxiv_id":null,"evidence_quote":"Introduces the Witten index which the paper adapts to a partial Witten index through the partial trace."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior work on the Witten index of massless $(d+1)$-Dirac-Schrödinger operators whose example and index formula are extended to full spectral shift functions."},{"cited_title":"Callias, Axial anomalies and index theorems on open spaces , Commun","cited_arxiv_id":null,"evidence_quote":"Original Callias index theorem that the paper's non-Fredholm index formula generalizes."}],"review_version":1}