{"id":"d2e20f81-174f-4328-8d66-df88b1b2246b","arxiv_id":"2607.18546","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A symmetry-aware variational ansatz based on the Holstein-Primakoff transformation simulates the finite-size Dicke model on qubit and qubit-boson hardware, demonstrated on trapped ions.","lead":"Researchers built a variational quantum eigensolver framework to simulate the finite-size Dicke model — a collective light-matter system — on both qubit-only and hybrid qubit-boson quantum computers. They demonstrate the approach on trapped-ion hardware and introduce a bosonic-variational version that uses fewer qubits.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unquantified linearization error in the spin-Dicke mapping leaves open whether finite-M digital simulations actually reproduce the Dicke model above λ_c.","rationale":"The reader's weakest assumption identifies the same gap: the linearized inverse Holstein-Primakoff mapping (Eq. 7) neglects higher-order terms, and the error relative to the actual Dicke model is not quantified for the hardware-relevant parameters. This is the most load-bearing concern because the central abstract claim—reproducing the Dicke model's critical behavior on digital processors—requires that the spin-Dicke Hamiltonian be a faithful representation of the Dicke model for the finite M used in the VQE and QPU experiments. The paper's evidence consists of comparisons between the spin-Dicke VQE results and exact diagonalization of the spin-Dicke Hamiltonian, not the Dicke model. The only direct link to the Dicke model is the thermodynamic-limit formula (Eq. 9), which is not derived and applies only as N→∞ with N/M→0, a limit far from the small systems tested. The test I propose would directly quantify the model error and settle whether the finite-M results are indicative of Dicke behavior. This is an addressable gap rather than a fundamental flaw: the variational method itself is sound, and the paper includes a systematic convergence study in M (Fig. 1b), suggesting the framework is valid in the appropriate limit. Therefore the existing CONDITIONAL verdict is appropriate, and no change is needed.","tokens_in":33002,"tokens_out":7007,"duration_ms":72976,"concrete_test":"For each (N,M) pair used in the paper (notably N=8, M=8 and N=4, M=4,6,8), compute the ground-state energy and ⟨a†a⟩/N of the exact Dicke Hamiltonian truncated to M+1 Fock states, and compare with the spin-Dicke model (Eq. 8) for λ/λ_c from 0 to 3. Quantify the maximum difference in energy and order parameter, e.g., |E_SD − E_Dicke|/N and |⟨a†a⟩_SD − ⟨a†a⟩_Dicke|/N. If the relative difference exceeds a few percent in the superradiant regime, the digital demonstration is not a faithful Dicke simulation. An alternative check is to include the next-order term from Eq. (6) in the Hamiltonian and verify whether the VQE results change appreciably.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim requires that the spin-Dicke model (Eq. 8) faithfully approximates the Dicke model for the parameters used on hardware (e.g., N=M=8, N=4 with M=4–8, N=1 with M=35). The mapping is obtained by dropping all terms beyond first order in the expansion of √(2s − (s+S_z)) (Eq. 6), replacing a ≈ S₋/√(2s) (Eq. 7). The error is controlled by the smallness of (s+S_z)/(2s) = a†a/M. For the superradiant phase (λ>λ_c), the Dicke ground state has ⟨a†a⟩ ∝ N. With M=N in the hardware demonstrations, a†a/M can be O(1) in this regime, so the expansion parameter is not small. The paper offers only a probabilistic bound on the tail probability of a†a/M and does not evaluate it for the specific states/parameters. Figure 1(b) shows a visible deviation between spin-Dicke and Dicke for N=10, M=10 at large λ; for M=8 the deviation is likely larger. The VQE and QPU results in Figs. 3–8 are compared only to the spin-Dicke model, not to the Dicke model. If the finite-M error is significant, the digital simulations solve a different (LMG-type) Hamiltonian, and the claim of reproducing the Dicke model's critical behavior is supported only in the unphysical limit M→∞, not for the demonstrated devices.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a variational quantum eigensolver (VQE) framework for the finite-size Dicke model. For fully digital qubit hardware, the bosonic mode is mapped via a linearized inverse Holstein–Primakoff transformation to M qubits, yielding a spin-Dicke Hamiltonian (Eq. (8)). A parity-, time-reversal-, and exchange-symmetric ansatz is constructed (Eqs. (35) and (36)) and benchmarked against exact diagonalization for N=4–10 and M=4–35. Classically optimized variational states are also executed on a trapped-ion QPU for N=1, M=35 and N=M=8. In addition, a hybrid qubit-boson ansatz is introduced (Eq. (37)) with an analytic energy functional (Eq. (38)). The central claim is that the resulting model reproduces the critical behavior of the Dicke model in the appropriate large-spin limit while remaining implementable on near-term devices.","tokens_in":33356,"tokens_out":22528,"duration_ms":215527,"significance":"The paper has genuine strengths: the analytic energy functionals in Eqs. (18), (38), and (C2) are correct and useful; the statevector VQE results match exact diagonalization for the spin-Dicke model; the trapped-ion demonstrations are a valuable proof of principle; and the symmetry analysis that reduces the ansatz to a single variational angle for N≤4 is elegant. However, the central claim that the framework simulates the Dicke model itself is undercut by two load-bearing issues: the thermodynamic-limit formula Eq. (9) is incorrect as written, and the error incurred by truncating the inverse Holstein–Primakoff expansion is not quantified for the hardware parameters used. Both issues are fixable, but they are central to the paper's main assertion.","major_comments":[{"comment":"The thermodynamic-limit formula is asserted without derivation and is incorrect as printed. For M=N, ω=ω0=1, and λ/λc=2, the radicand is 1 − 4(1 − 1/16) = −11/4, so Eq. (9) gives an imaginary number, while the spin-Dicke ground state is real (Fig. 1(b) shows ⟨a†a⟩/N ≈ 0.25). The correct two-spin mean-field result is ⟨a†a⟩/N = M/(2N)[1 − sqrt((1 + Nω0²/(4Mλ²))/(1 + 4Nλ²/(Mω²)))] for λ > λc, which reduces to the Dicke formula when N/M→0 and yields the slope in Eq. (10). Eq. (9) should be replaced and its derivation supplied; as written it undermines the thermodynamic-limit claim.","section":"Section III, Eq. (9)"},{"comment":"The linearization a ≈ S₋/√(2s) discards all higher-order terms in Eq. (6). The paper controls this only by a probabilistic tail bound and does not quantify the actual error between the spin-Dicke and Dicke Hamiltonians for the parameters demonstrated on hardware (N=M=8 and N=1, M=35). All VQE and QPU comparisons are against exact diagonalization of the spin-Dicke model, not the Dicke model. For N=M=8 and λ>λc, a†a/M is not small (≈0.25 at λ/λc=2 and up to 0.5 in the strong-coupling limit), and Fig. 1(b) shows visible spin-Dicke/Dicke deviations for N=M=10 at large λ. Please add a quantitative convergence study (e.g., energy or fidelity vs M for N=8) or temper the claim that the demonstrated digital simulations reproduce the Dicke model itself.","section":"Section III, Eq. (7), and Figs. 3–8"}],"minor_comments":[{"comment":"The sentence 'For λ=0, the expectation value can reach the ground-state energy if cosθ=−1' is inconsistent with Eq. (18); minimizing Eq. (18) at λ=0 gives cosθ=1. Please fix this typo.","section":"Section V.B, text before Eq. (18)"},{"comment":"Typo: 'orange dashed linies' should be 'orange dashed lines'.","section":"Figure 5 caption"},{"comment":"The stated Hilbert-space dimension ((2s+1)·(2j+1)^{N−1})² appears incorrect; for N spins-j plus one spin-s the dimension is (2s+1)(2j+1)^N. Please check the resource estimate.","section":"Appendix A"},{"comment":"Ref. [41] is referred to as 'the manuscript in question'; for clarity, use 'Ref. [41]' instead.","section":"Introduction, Ref. [41]"}],"recommendation":"major_revision","confidential_remarks":"The paper is suitable for this journal if the authors correct Eq. (9) and add a quantitative convergence study of the spin-Dicke to Dicke mapping for the demonstrated parameters. The incorrect thermodynamic-limit formula and the unquantified linearization error are the main barriers; the rest of the methodology and the QPU demonstrations are solid."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Take this paper as a useful variational method paper rather than a definitive Dicke-model simulation. The genuinely new pieces are the exchange-symmetric collective MS ansatz with recursively fixed angles, the closed-form energy functionals (Eqs. 18, 38, and C2), and the hybrid qubit-boson displacement ansatz. These are derived cleanly and checked against exact diagonalization for N=4-10, M=4-35; the statevector VQE agreement is solid. The trapped-ion data for N=1, M=35 and N=M=8 are honest proof-of-principle runs, and the paper openly says the hardware is noisy.\n\nThe main soft spot is exactly what the stress-test note says: the linearized inverse Holstein-Primakoff mapping a≈S-/√(2s) is controlled only by the smallness of a†a/M, and in the superradiant phase at fixed M=N this quantity can be O(1). The paper gives a probabilistic bound but never evaluates it for the specific states in Figs. 3-8. So the digital results are really for the spin-Dicke model, not the Dicke model, and the abstract's 'reproduces characteristic critical behavior' should be read as 'in the large-M limit,' which the paper does state. The fix is straightforward: for each (N, M, λ) shown, report ⟨a†a⟩/M and a conservative estimate of the truncation error, or add a Tomonaga-Luttinger-style correction. Without that, the reader cannot tell whether the N=M=8 QPU demo is simulating a model with significant deviations from the Dicke Hamiltonian.\n\nTwo lesser issues. Eq. (9) is asserted without derivation; it is a mean-field result and probably right, but it should be derived or referenced properly. And the 'scalable' claim outruns the evidence: the recursively determined angles are clever, but the paper only demonstrates systems up to N=M=10 classically and N,M=8 on hardware. That is small, not scalable. QPU points also lack error bars, which matters because the measurements are from a small number of shots.\n\nOn the positive side, the analytic work is reproducible, the comparison to exact diagonalization is honest, and the paper does not oversell the symmetry-preserving ansatz's ability to capture symmetry-broken states. The citation pattern seems fair. No circularity: the ansatz is fitted to the spin-Dicke Hamiltonian, not to the Dicke target.\n\nThis is a solid methods paper for people doing variational quantum simulation of collective light-matter systems. It deserves a serious referee, not a desk reject, but the revision should quantify the linearization error and tone down the scalability language. I'd cite it if I were working in this niche.","headline":"Solid variational method paper for the spin-Dicke model, but the mapping to the actual Dicke model is only controlled in the large-M limit, and the paper never quantifies the error for the finite-M hardware demos.","tokens_in":33884,"tokens_out":2792,"would_cite":true,"duration_ms":32054,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A truncated spin-mapped version of the Dicke model, solved with a symmetry-preserving variational ansatz, reproduces the finite-size critical behavior of the full model on digital and trapped-ion quantum computers.","keywords":["Dicke model","superradiant phase transition","variational quantum eigensolver","Holstein-Primakoff transformation","symmetry-preserving ansatz","trapped-ion quantum computer","hybrid qubit-boson simulation","finite-size critical behavior"],"falsifier":"Compute the full Dicke model ground state (using exchange-symmetry reduced diagonalization) for N=8, M=8 at λ=2λ_c, and compare its energy and ⟨a†a⟩/N with the spin-Dicke variational result reported in the paper; if the difference exceeds the variational error of ~10^-2, the mapping truncation — not the ansatz — is the dominant error source, falsifying the claim that the framework simulates the Dicke model at these hardware parameters.","tokens_in":32910,"feed_emoji":"⚛️","tokens_out":7968,"duration_ms":75575,"temperature":0.7,"pith_summary":"The paper aims to show that the Dicke model — the standard quantum-optics description of many two-level atoms coupled to one cavity mode — can be simulated on modest quantum hardware by replacing the bosonic mode with a finite spin via the leading-order inverse Holstein-Primakoff mapping. The resulting 'spin-Dicke model' keeps the superradiant phase transition at the same critical coupling as the full model when the spin size is large enough. Using a variational ansatz that preserves parity and exchange symmetries, the authors match exact diagonalization for ground and excited states across a wide coupling range, with systems up to N=10 atoms and M=35 cavity qubits. The circuits were run on a trapped-ion processor, and a hybrid qubit-boson variant reduces qubit count by keeping the cavity mode continuous.","feed_headline":"A symmetry-preserving circuit reproduces the Dicke phase transition","feed_subtitle":"Variational eigensolver matches exact results up to 10 atoms and 35 cavity qubits, with trapped-ion runs confirming the ansatz.","key_machinery":"The key object is the linearized inverse Holstein-Primakoff mapping a → S_-/√(2s), which truncates the infinite bosonic Hilbert space to a spin-s register (s=M/2) and produces the spin-Dicke Hamiltonian. The variational workhorse is a collective parity-preserving gate e^{-iθ X_a Y_c/2} applied to a maximally entangled atomic state in the X basis, whose layer-by-layer construction enforces exchange symmetry and reduces the full ansatz to a single variational angle.","core_discovery":"The central discovery is that the finite-size Dicke Hamiltonian can be mapped, via the first-order inverse Holstein-Primakoff transformation a → S_-/√(2s), to an all-qubit 'spin-Dicke' Hamiltonian that, in the large-spin limit, shows a second-order phase transition at the same λ_c = √(ωω0)/2 as the full model. The authors' symmetry-preserving variational state — a collective gate e^{-iθ X_a Y_c/2} applied to an exchange-symmetric atomic state — yields energies matching exact diagonalization to order 10^-2 across the coupling range, and runs on trapped-ion hardware up to N=8, M=8. A hybrid variant using the bosonic mode directly (e^{-ix X_a P}) matches the exact N=10 ground state within 4%.","pith_inferences":["The paper demonstrates the spin-Dicke approximation only against exact diagonalization of the spin-Dicke Hamiltonian, not the full Dicke Hamiltonian; an obvious next check is to benchmark the mapping error itself against exchange-symmetric exact Dicke solutions for the same (N,M) pairs.","If the first-order mapping error remains small for larger N and off-resonant couplings, the same symmetry-preserving ansatz structure transfers almost unchanged to other collective spin-boson models, which the paper mentions but does not test.","The infidelity-minimization procedure that fixes the atomic angles (Eq. 33) suggests a quantitative scalability criterion: if the minimized infidelity grows with N, the single-angle ansatz must be deepened — this is a testable prediction of the paper's own construction."],"forward_implications":["The spin-Dicke model is a systematically improvable approximation: raising M reduces the mapping error, and the critical coupling stays fixed at λ_c for every N/M ratio.","Near-term all-to-all devices like trapped-ion quantum processors can study finite-size precursors of the superradiant transition using as few as N + M qubits, demonstrated here at N=8, M=8.","Excited states are obtained without deflation or penalty terms, simply by initializing within a chosen parity sector, simplifying excited-state variational quantum eigensolver calculations.","The hybrid qubit-boson ansatz lowers qubit count and circuit depth enough that near-critical-point studies may be possible even when only a small number of bosons can be supported by the hardware."],"fun_headline_variants":["Dicke phase transition reproduced by variational quantum circuits","Variational circuits mimic light-matter phase transition","Hybrid qubit-boson approach simulates Dicke model","Symmetry-aware ansatz simulates Dicke model on quantum hardware","Quantum algorithms capture Dicke model's critical behavior"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that keeping only the first term of the inverse Holstein-Primakoff expansion, a ≈ S_-/√(2s), produces negligible error for the low-lying eigenstates at the system sizes and couplings used; if that error is large, the digital simulations would solve a different collective spin model rather than the Dicke model.","fun_headline_variants_meta":{"raw":{"variants":["Dicke phase transition reproduced by variational quantum circuits","Variational circuits mimic light-matter phase transition","Hybrid qubit-boson approach simulates Dicke model","Symmetry-aware ansatz simulates Dicke model on quantum hardware","Quantum algorithms capture Dicke model's critical behavior"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000425,"raw_usage":{"total_tokens":2024,"prompt_tokens":761,"completion_tokens":1263,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":505,"completion_tokens_details":{"reasoning_tokens":1194}},"tokens_in":505,"tokens_out":1263,"duration_ms":10667,"temperature":1.0,"reasoning_tokens":1194,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T15:05:18.184573+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full Dicke model ground state (using exchange-symmetry reduced diagonalization) for N=8, M=8 at λ=2λ_c, and compare its energy and ⟨a†a⟩/N with the spin-Dicke variational result reported in the paper; if the difference exceeds the variational error of ~10^-2, the mapping truncation — not the ansatz — is the dominant error source, falsifying the claim that the framework simulates the Dicke model at these hardware parameters.","supporting_citations":[],"review_version":1}