{"id":"c52b65d9-d34e-4c5c-b42c-8e16a8011b92","arxiv_id":"2607.22378","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For normal couplings, k-photon light–matter Hamiltonians are self-adjoint for k≤2 and, for k≥3, have deficiency spaces isomorphic to k copies of the coupling's nonzero range.","lead":"This paper proves when a family of k-photon light–matter Hamiltonians, where a field and a quantum system exchange k excitations at once, has a well-defined quantum dynamics. For k≥3 and ordinary (normal) couplings the Hamiltonian is not self-adjoint, so extra boundary conditions are needed; the paper counts and classifies those choices.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem's advertised 'normal nonzero' scope is not proven: Proposition 3.11 requires a bounded inverse on (KerΣ)^⊥, which fails for e.g. compact diagonal Σ with 0 in the essential spectrum.","rationale":"The reader's weakest-assumption identification is exactly the bounded-inverse/spectral-gap condition on Σ1. This is the most load-bearing issue because it is the only place where the proof's machinery — block Jacobi operators with boundedly invertible coefficients and Świderski's theorem — can fail to apply to a normal nonzero Σ explicitly allowed in the abstract. The core proof for the stated Theorem 1 (with bounded inverse) appears internally consistent: the decomposition into block Jacobi operators is correct, the Carleman condition handles k≤2, and the unitary transformation using the polar decomposition of normal Σ correctly produces ̃A_n = β_n Π, ̃B_n = ω(...)1 + (U*)^n H_mat U^n, making the Świderski limits T=Q=0, R=1, C=Π/||Π|| and giving a strictly positive form. No algebraic error was found in the main theorem. The gap is one of scope: the abstract claims a theorem for all nonzero normal Σ while the proof requires an extra spectral-gap condition. The concrete test with diag(1/n) would settle whether the conclusion nonetheless extends; if it does, the paper still needs a direct-integral argument to support the abstract. I therefore keep the reader's CONDITIONAL verdict unchanged.","tokens_in":19858,"tokens_out":15585,"duration_ms":142875,"concrete_test":"Take ℋ=ℓ^2(ℕ), Σ=diag(1/n)_{n≥1}, k=3, H_mat=0. Compute the deficiency indices of H = ⊕_{n≥1}(ωa*a + (1/n)(a*)^3 + (1/n)a^3) using the known result that each scalar summand has deficiency (3,3). If the deficiency space is not isomorphic to ⊕_{m=0}^{2}(KerΣ)^⊥, then the bounded-inverse assumption is essential and the abstract claim is false. If it is isomorphic, the missing bounded-inverse case needs an explicit direct-integral proof before the advertised 'normal nonzero' result can be accepted as proven.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract and the opening of Theorem 1 state the result for every normal nonzero Σ, but the proof of Proposition 3.11 assumes that Σ1 = Σ|(KerΣ)^⊥ has a bounded inverse. This is not automatic for bounded normal operators: on ℓ^2, Σ=diag(1/n) is normal, nonzero, injective, yet Σ1 has no bounded inverse. Under such Σ, the block Jacobi coefficients A_n^{(m)} = β_{m+nk}Σ* in Proposition 3.3 are not boundedly invertible, so Definition 2.1 and Świderski's Theorem 2.6 cannot be applied to the nonzero part. The proof therefore stops at the point where the abstract's advertised generality begins. Remark 3.12 explicitly acknowledges that a direct-integral argument would be needed and does not supply it. The conclusion may still be true: each nonzero scalar fiber ωa*a + λ(a*)^k + \\barλ a^k has deficiency (k,k), suggesting total deficiency k·dim((KerΣ)^⊥). But that is a separate argument, not the one in the paper. Thus the paper proves the classification only under an unadvertised spectral-gap condition, and the abstract overstates what is established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a class of light–matter Hamiltonians of the form H = H_mat⊗1 + 1⊗ωa*a + Σ⊗(a*)^k + Σ*⊗a^k on ℋ⊗L^2(ℝ), where the matter operator Σ is bounded. The main result (Theorem 1, supported by Propositions 3.5, 3.9, 3.11 and Corollary 3.10) states that, for normal Σ with bounded inverse on (KerΣ)^⊥, the operator is self-adjoint for k≤2; for k≥3 it is not self-adjoint and its deficiency spaces are isomorphic to ⊕_{m=0}^{k−1}(KerΣ)^⊥. In finite-dimensional matter systems all self-adjoint extensions are parametrized by kd'×kd' unitaries and have purely discrete spectrum. The proof expands in Fock states to obtain a direct sum of block Jacobi operators, uses the Carleman criterion for k≤2, and for k≥3 applies a unitary transformation built from the polar decomposition of Σ so that Świderski's theorem applies. The paper closes with applications to the k-photon Rabi and Dicke models and a non-normal Jaynes–Cummings counterexample.","tokens_in":20128,"tokens_out":8336,"duration_ms":83733,"significance":"If fully established for the advertised generality, this would be a valuable contribution: it gives a unified block-Jacobi treatment of multiphoton light–matter Hamiltonians, an explicit verification of the hypotheses of Świderski's theorem, and concrete, falsifiable predictions for the deficiency indices and spectral nature of extensions. The method is elegant and does not rely on fitted parameters or model-specific Bargmann computations. However, the proof as written requires a spectral-gap condition on Σ that is not mentioned in the abstract; the gap between the advertised and the proved statement is the main issue.","major_comments":[{"comment":"The abstract states 'When Σ is normal and nonzero, we prove ...' and the opening of Section 1 repeats this scope. The proof, however, only covers Σ for which Σ₁ = Σ|(KerΣ)^⊥ has a bounded inverse, as assumed in Proposition 3.11. This condition is not automatic for bounded normal operators: on ℓ², Σ = diag(1/n) is normal, nonzero, and injective, but Σ₁^{-1} is unbounded because 0 is in the essential spectrum. In that case the block Jacobi coefficients A_n^{(m)} = β_{m+nk}Σ* in Proposition 3.3 lack bounded inverses, so Definition 2.1 and Świderski's Theorem 2.6 cannot be applied to the nonzero part. Remark 3.12 explicitly says that a direct-integral argument 'would be needed' to handle general Σ, but no such argument is supplied. The formal Theorem 1 in the introduction already includes the bounded-inverse condition, so the abstract overstates what is established. This is load-bearing: the","section":"Abstract; §3.2, Proposition 3.11; Remark 3.12"},{"comment":"The sentence 'Therefore, Proposition 3.11 applies to all operators from Definition 3.1: if Σ is normal and nonzero, H is self-adjoint if and only if k≤2' is correct only in finite dimension, where the restriction of Σ to (KerΣ)^⊥ is automatically invertible. In infinite dimension it repeats the overclaim discussed above. Since the sentence follows a finite-dimensional remark it may be intended to be read in that context, but as written it is misleading and should be qualified explicitly.","section":"§3.2, final paragraph before Remark 3.12"}],"minor_comments":[{"comment":"In Eq. (3.64) the deficiency-space formula uses P_n^{(m)}(∓i). Since Lemma 2.3 states the formula with z*, the sign convention is plausible but slightly confusing. A one-line verification that P_n^{(m)}(∓i) matches the z* convention would improve readability.","section":"§2, Definition 2.2 and Theorem 3.9"},{"comment":"The claim that replacing the domain by 𝒮(ℝ) or by 𝒟((a*a)^{m/2}) gives the same closure is stated without proof. It is likely true by relative boundedness, but a short justification or reference would be helpful.","section":"Remark 3.2"},{"comment":"The Jaynes–Cummings example convincingly shows that normality cannot simply be dropped from the hypotheses. It would be useful to state explicitly that this example does not address the additional bounded-inverse condition, so the 'optimality' claim is only about the normality assumption within the class treated in the paper.","section":"§4, Example 4.3"}],"recommendation":"major_revision","confidential_remarks":"The technical core is sound under the restrictive hypothesis stated in Theorem 1: the application of Świderski's theorem is careful and the finite-dimensional applications are correct. The main problem is a scope mismatch: the abstract overclaims to all nonzero normal Σ while the proof requires a spectral gap. The missing general case is explicitly identified in Remark 3.12 and is likely tractable by direct integrals, so the paper is worth a major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a well-built paper on the right problem. It proves that for H = H_mat⊗1 + 1⊗ωa*a + Σ⊗(a*)^k + Σ*⊗a^k, with Σ normal and having bounded inverse on (KerΣ)^⊥, k≤2 gives self-adjointness and k≥3 gives deficiency indices k·dim((KerΣ)^⊥), parametrizes the extensions, and shows discrete spectrum in finite dimension. The block Jacobi decomposition from [20] plus the unitary transform based on the polar decomposition is the real work, and the verification of Świderski's conditions is clean. I checked the asymptotics in Lemma 3.7 and the form of the transformed coefficients in Lemma 3.6; both hold. The treatment of the k-photon Rabi and Dicke models is accurate, and the Jaynes–Cummings counterexample makes the normality assumption genuinely necessary.\n\nThe soft spot is exactly where the stress-test note lands. The abstract says \"Σ normal and nonzero\"; Theorem 1 in the introduction includes the bounded-inverse condition on (KerΣ)^⊥, and the detailed Proposition 3.11 relies on it. That condition is not automatic for a bounded normal Σ: take Σ = diag(1/n) on ℓ². It is normal, nonzero, and injective, but its restriction to (KerΣ)^⊥ has no bounded inverse. In that case the block Jacobi coefficients A_n = β_n Σ* are not boundedly invertible and Definition 2.1/Świderski's theorem do not apply to the full problem. So the paper proves the classification only under an unadvertised spectral-gap assumption. The authors are aware; Remark 3.12 sketches a direct-integral route but does not carry it out. The result may well be true in that generality, but this manuscript does not establish it.\n\nHow much does this matter? For finite-dimensional matter systems—which covers the physical examples in Section 4—the condition is automatic and the paper is fully rigorous. For infinite-dimensional matter with 0 in the essential spectrum of Σ, it is a genuine gap between the advertised theorem and the proof. That should be fixed in revision, either by stating the assumption in the abstract and main theorem or by proving the general case.\n\nThe paper is honest, clearly written, and the mathematics under the stated assumptions is sound. The citations, including the self-citations, are used correctly. It deserves a serious referee. I would send it to review and tell the authors to align the abstract with the theorem. For a reading group it is a useful example of block Jacobi techniques, though not an urgent one.","headline":"Solid extension theory for k-photon models under a spectral-gap condition on Sigma; the abstract, though, promises more than the proof delivers for infinite-dimensional normal Sigma.","tokens_in":20611,"tokens_out":2424,"would_cite":true,"duration_ms":24790,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["47B36","47B25","81Q10","81Q12","46N50"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that for k-photon light-matter Hamiltonians with a normal nonzero coupling, self-adjointness holds exactly for k≤2, while k≥3 forces a family of self-adjoint extensions, all with discrete spectra in finite dimension.","keywords":["self-adjoint extensions","deficiency indices","block Jacobi operators","multiphoton interactions","quantum Rabi model","Dicke model","normal operators","light-matter Hamiltonians"],"falsifier":"Take the k-photon Rabi model at k = 3 and solve the generalized eigenvalue equation directly: the paper predicts that the space of square-integrable solutions at a nonreal spectral parameter has dimension 6. A different count would refute Theorem 1. A second check: choose Σ as a diagonal normal operator on ℓ² with eigenvalues 1/n and verify whether the deficiency indices for k = 3 still equal k times the dimension of (KerΣ)^⊥.","tokens_in":19737,"feed_emoji":"⚛️","tokens_out":9539,"duration_ms":84088,"temperature":0.7,"pith_summary":"Multiphoton light-matter interactions, where a bosonic mode exchanges k excitations at a time with a quantum system, are becoming experimentally realistic. This paper asks whether the natural Hamiltonian defines a unique quantum dynamics, and answers with a clean threshold: for k≤2 it is self-adjoint, for k≥3 it is not, whenever the coupling operator is normal and nonzero (with a bounded inverse on the complement of its kernel). The missing self-adjointness is measured explicitly: the deficiency spaces are k copies of the coupling range, so in finite dimension one must choose a kd'×kd' unitary to specify boundary conditions. Every such choice still has purely discrete spectrum, so the ambiguity is real but tame. A non-normal counterpart, the k-photon Jaynes-Cummings model, stays self-adjoint for all k, showing the condition is sharp.","feed_headline":"At three or more photons, Hamiltonians lose unique dynamics","feed_subtitle":"Normal couplings are self-adjoint for k≤2; for k≥3, boundary data pick among infinitely many valid Hamiltonians.","key_machinery":"The load-bearing mechanism is a decomposition of H into finitely many block Jacobi operators—infinite tridiagonal operator matrices acting on sequences of matter states, with diagonal blocks ω(m+rk)1 + H_mat and off-diagonal blocks β_{m+rk}Σ*. For k ≤ 2 these blocks grow slowly enough that a classical summability criterion proves self-adjointness. For k ≥ 3, the paper conjugates each block by powers of the unitary part of the polar decomposition Σ = UΠ; normality makes U and Π commute, turning the off-diagonal coefficients into positive operators βΠ with bounded inverses. The transformed operators satisfy a known theorem on block Jacobi operators that yields maximal indeterminacy, i.e., defi","core_discovery":"The central claim, stated as Theorem 1, is a dichotomy for H = H_mat ⊗ 1 + 1 ⊗ ωa*a + Σ ⊗ (a*)^k + Σ* ⊗ a^k. If Σ is a bounded normal operator with nonzero values on the subspace orthogonal to its kernel, then H is essentially self-adjoint for k ≤ 2. For k ≥ 3, H is not self-adjoint: its two deficiency spaces are both isomorphic to the direct sum of k copies of (KerΣ)^⊥, and every self-adjoint extension is parameterized by a unitary between those deficiency spaces. When the matter system is finite-dimensional, the deficiency index is kd' with d' = dim(KerΣ)^⊥, the extensions are kd'×kd' unitaries, and the spectrum of every extension is purely discrete. The proof obtains this by decomposing H","pith_inferences":["This decomposition suggests the same block-Jacobi route can treat more general polynomial bosonic interactions; a natural test is whether subordinacy-type arguments predict absolutely continuous spectrum for particular self-adjoint extensions.","Because boundary conditions are unitaries, one could numerically compare two extensions of the k-photon Rabi model (k = 3) and look for experimentally distinguishable spectral phases, e.g., different low-lying eigenvalues.","The paper gives only a lower bound on deficiency indices for non-normal couplings sharing eigenvectors; checking whether this lower bound is sharp for diagonalizable but non-normal Σ would map how far beyond normality the counting extends.","The conserved-number mechanism in the Jaynes-Cummings model hints at a broader criterion: any multiphoton interaction admitting a block decomposition into finite-dimensional invariant subspaces will be self-adjoint, independent of k."],"forward_implications":["For k ≤ 2, every model in this class is self-adjoint on the minimal domain; no boundary data are needed.","For k ≥ 3 with finite-dimensional matter, all self-adjoint extensions have purely discrete spectrum, so any chosen boundary condition yields bound-state-only dynamics.","States in the kernel of Σ contribute no deficiency; the number of missing boundary conditions is controlled entirely by dim(KerΣ)^⊥ and k.","If Σ is normal and invertible, each deficiency space is a copy of the full matter Hilbert space, so the extension parameter is a unitary acting on k copies of that space.","The k-photon Jaynes-Cummings model remains self-adjoint for all k because a conserved excitation number decomposes H into finite-dimensional blocks—showing the normality assumption is essential."],"fun_headline_variants":["Normal coupling: multiphoton Hamiltonian self-adjoint only for k≤2","k≥3 with normal coupling: Hamiltonian not unique, multiple extensions","Three or more photons break uniqueness for normal coupling","For normal light-matter coupling, k≥3 loses Hamiltonian uniqueness"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The proof's load-bearing premise is that the normal coupling Σ has a bounded inverse on (KerΣ)^⊥, so its nonzero spectrum stays away from zero; if Σ's spectrum accumulates at 0, the block Jacobi coefficients lose the invertibility that the indeterminacy argument requires.","fun_headline_variants_meta":{"raw":{"variants":["Normal coupling: multiphoton Hamiltonian self-adjoint only for k≤2","k≥3 with normal coupling: Hamiltonian not unique, multiple extensions","Three or more photons break uniqueness for normal coupling","For normal light-matter coupling, k≥3 loses Hamiltonian uniqueness"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001807,"raw_usage":{"total_tokens":6992,"prompt_tokens":823,"completion_tokens":6169,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":567,"completion_tokens_details":{"reasoning_tokens":6095}},"tokens_in":567,"tokens_out":6169,"duration_ms":43333,"temperature":1.0,"reasoning_tokens":6095,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T04:58:04.073537+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the k-photon Rabi model at k = 3 and solve the generalized eigenvalue equation directly: the paper predicts that the space of square-integrable solutions at a nonreal spectral parameter has dimension 6. A different count would refute Theorem 1. A second check: choose Σ as a diagonal normal operator on ℓ² with eigenvalues 1/n and verify whether the deficiency indices for k = 3 still equal k times the dimension of (KerΣ)^⊥.","supporting_citations":[],"review_version":1}