{"id":"6b3c1256-e81f-457a-a69e-9f9854c7ba1c","arxiv_id":"2607.22850","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"ViBra is a validated open-source VSCF/VCI program for anharmonic vibrational spectra that can take externally generated, including quantum-sampled, configuration lists as seeds and refine them classically.","lead":"A new open-source program, ViBra, computes anharmonic molecular vibrations by combining VSCF/VCI methods with configuration selection and symmetry blocking, and it can accept configuration lists generated by quantum sampling algorithms. The paper validates the code against reference calculations and demonstrates a simulated quantum-classical workflow for water.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Semiquartic-only benchmarks leave full-quartic coupling path unvalidated; S-VCI performance on complete quartic PES is untested.","rationale":"The reader's weakest assumption noted the semiquartic force field, but framed it as a limitation on physical accuracy of the PES. My concern is more targeted: the semiquartic restriction means the code's handling of all-distinct quartic coupling terms—terms explicitly supported by the implementation and implied by the abstract's 'quartic force field'—is never validated. This is a correctness risk, not merely a physical-accuracy gap. The VSCF and full VCI modules are validated against Crystal23 for water, which likely uses a full quartic PES, but the SA-VCI and S-VCI modules (and the quantum-seeded H2O demonstration) are only tested on semiquartic PES. Since SA-VCI is mathematically exact, the risk is mainly for S-VCI selection and the Hamiltonian elements involving four distinct modes. The S-VCI accuracy claim is based on a single molecule (ethylene) with a semiquartic PES, so extrapolating to full quartic is unsupported. This does not invalidate the paper's core implementation claims for the tested inputs, but it warrants a conditional acceptance: the authors should either add a full-quartic test or explicitly scope the method to semiquartic force fields as provided by ORCA VPT2. The paper is otherwise well-structured with reproducible data and transparent disclaimers about simulator-based quantum work.","tokens_in":24059,"tokens_out":12238,"duration_ms":117925,"concrete_test":"Run ViBra on a molecule with a complete quartic force field containing non-negligible all-distinct terms φ_ijkl (e.g., H2O or formaldehyde at a level of theory that yields such terms). (1) Compare full VCI energies to an independent implementation (e.g., pyVCI or VHCI) on the same PES; (2) run S-VCI with increasing MAXSCI and compare to full VCI. If energies match and S-VCI errors stay within the claimed 1–5 cm⁻¹ range, the concern is resolved. Alternatively, inspect the ORCA .vpt2 files provided in the Zenodo archive to check whether any contain all-distinct quartic terms; if none do, the benchmark set does not cover the code's advertised full-quartic capability.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's benchmarks for SA-VCI, S-VCI, and the H2O quantum-seeded workflow all employ semiquartic force fields—defined as cubic terms plus quartic contributions with at least two repeated mode indices. All-distinct quartic terms φ_ijkl (i<j<k<l) are therefore absent from every validation, even though the method is presented as starting from a full quartic force field (Eq. 2) and the Hamiltonian construction (Eq. 19, sparse pair list with d≤4, inverted potential index) explicitly supports such terms. As a result, the central claim that ViBra is a correct VCI/S-VCI engine for quartic force fields is not fully established: a bug in the all-distinct quartic coupling path would go undetected, and the demonstrated S-VCI accuracy (within ~1–5 cm⁻¹ of full VCI in Table 4) could be specific to the semiquartic PES. The paper also does not clarify whether ORCA's .vpt2 format typically supplies full quartic constants; if not, the full-quartic capability is unused and the validation covers only the actual input domain, which would mitigate the concern but still leaves an unexercised advertised feature.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents ViBra, an open-source Fortran/Python package for anharmonic vibrational spectroscopy built on a quartic Taylor expansion of the potential energy surface in normal coordinates. The workflow combines VSCF with full, selected (S-VCI), and symmetry-adapted (SA-VCI) configuration interaction, evaluates infrared intensities including electrical anharmonicity, and provides a GUI/spectral viewer. The authors also describe an interface that accepts externally generated configuration lists, including a quantum-sampling proof-of-concept for water using a boson-to-qubit encoding, Trotterized time evolution, and simulated measurement bitstrings. Validation is four-tiered: VSCF and VCI against Crystal23 water data (0.001-0.01 cm-1 agreement), SA-VCI against full VCI (numerical equivalence), S-VCI variational convergence, and harmonic intensities against ORCA. The central claims are that ViBra is numerically correct, that S-VCI reproduces full VCI to within a few cm-1 at reduced cost, and that the quantum-seeded workflow works as an interface. The quantum results explicitly disclaim hardware execution and quantum advantage.","tokens_in":24373,"tokens_out":4121,"duration_ms":45967,"significance":"If the validation gaps are closed, ViBra would be a useful, reproducible open-source contribution to anharmonic vibrational spectroscopy: it provides a GUI, a selected-CI scheme based on Epstein-Nesbet PT2, symmetry-adapted block diagonalization for Abelian groups, and a natural interface for sample-based quantum-seeded VCI. The paper's strengths include open code (GitHub), a Zenodo data archive with inputs and outputs, explicit disclosure that the quantum workflow is simulator-only, and a four-tier validation strategy that goes beyond self-consistency by benchmarking against Crystal23 and ORCA. The significance is moderate: the core numerical engine appears sound, but the generality of the S-VCI and full-quartic-force-field claims is not yet established by the benchmarks provided.","major_comments":[{"comment":"All validation runs (SA-VCI, S-VCI, and the H2O quantum workflow) use semiquartic force fields, defined in Step 3 as cubic constants plus quartic terms with at least two repeated mode indices. Terms with all-distinct quartic indices (φ_ijkl, i<j<k<l), which Eq. (2) explicitly includes and Eq. (19) claims to implement, are never exercised. A bug in that code path would go undetected, and the demonstrated S-VCI accuracy on semiquartic PES does not transfer automatically. Please add at least one benchmark with a full quartic force field—e.g., water with complete φ_ijkl compared against Crystal23 or another reference—and clarify whether ORCA .vpt2 output typically contains all-distinct quartic constants; if not, state that this advertised capability is dormant.","section":"§2.4, Eq. (19); §5.1, Step 3; §6.1"},{"comment":"The claim that S-VCI reproduces full VCI to within 1-5 cm-1 is supported only by the ethylene example at NQUANT=6. Step 4 reports monotonic variational lowering for the whole test set, but monotonicity alone does not quantify accuracy; no S-VCI versus full-VCI energy differences are given for water, HFCO, N2H4, or trans-N2H2. This limits the generality of the S-VCI claim. Please include S-VCI/full-VCI error statistics for at least one additional molecule (e.g., water at NQUANT=6-8) or explicitly scope the accuracy claim to ethylene.","section":"§5.2, Table 4; §5.1, Step 4"},{"comment":"The paper never compares computed anharmonic frequencies or intensities to measured values for any molecule; the only experimental comparison is a qualitative ethylene overlay in Fig. 2. Cross-code agreement (Crystal23, ORCA) and internal consistency (SA-VCI vs full VCI) establish numerical correctness but not that the quartic-force-field workflow is 'physically faithful' in the spectroscopic sense claimed in Section 7. A quantitative comparison of at least a few fundamentals and overtones/combinations for water or ethylene against experimental gas-phase data (frequencies and, where available, relative intensities) would substantially strengthen the paper.","section":"§5.2, Fig. 2; §7"}],"minor_comments":[{"comment":"Table 4 states 'All calculations used 8 OpenMP threads,' but the discussion of SA-VCI scalability refers to 'all 16 threads' and thread saturation. Reconcile the thread count or clarify the hardware/thread configuration for each timing.","section":"§5.2, Table 4 versus text"},{"comment":"The term 'semiquartic' is introduced in Step 3 but should be defined in Section 2.1 where Eq. (2) presents the full quartic expansion, since the distinction is central to the validation scope.","section":"§2.1/§5.1"},{"comment":"In Figure 5 and the surrounding text, configuration labels such as (0,2,0) and (5,0,0) are understandable in context, but the mode ordering for water's three normal modes should be stated explicitly in the figure caption or text.","section":"§5.2/§6.1"},{"comment":"The intensity normalization uses Ecutoff with a default of 4500 cm-1; clarify in the text whether this is a display-only threshold and that absolute intensities require source modification, which is already mentioned but could be more prominent.","section":"Eq. (33)"}],"recommendation":"major_revision","confidential_remarks":"The reader's report recommends acceptance, but I see the semiquartic-only validation as a load-bearing gap: the advertised full-quartic capability and the general S-VCI accuracy claim are not tested on the full operator set. This is fixable with additional benchmarks, not a rejection. The manuscript is otherwise careful, transparent about its simulator-only quantum results, and well supported by data/code releases. Fit with the journal's scope is acceptable given the quantum-centric integration angle."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good read overall. ViBra is a serious, usable VSCF/VCI code, and the validation is genuinely careful. The genuinely new thing is the state-list interface: an external routine can dump a set of vibrational configurations into the code, and ViBra either diagonalizes the projected Hamiltonian directly or enlarges the space through EN-PT2 selection before rediagonalizing. That is a natural fit for SQD-style quantum sampling, and I don’t recall another VCI package offering it. The four-tier benchmark is appropriate: water VSCF/VCI against Crystal23 to 0.001–0.01 cm⁻¹, SA-VCI reproducing full VCI to ~10⁻² cm⁻¹ on five molecules, S-VCI converging monotonically, and harmonic intensities matching ORCA. Code and data are on GitHub/Zenodo, making the results reproducible; the GUI is a plus.\n\nThe real gap is the full-quartic coupling path. Every benchmark in Steps 3, 4, and 6 uses semiquartic force fields—cubic plus quartic terms with at least two repeated indices. The code's Hamiltonian construction and pair list explicitly support all-distinct quartic terms, but that path is never exercised. A bug there would go undetected, and the demonstrated S-VCI accuracy could be specific to the semiquartic PES. It is plausible that ORCA's .vpt2 format does not supply all-distinct quartic constants, which would soften the concern, but then the advertised full-quartic capability is unused. Either way, the paper should state this explicitly and scope its correctness claim to what was tested.\n\nSecond, experimental validation is weak: no anharmonic frequencies are compared to measured values—the only experiment is a qualitative ethylene overlay. That limits the spectroscopic significance of the claims, though it is a software-description paper, not a benchmark of force-field accuracy. Third, the H2O quantum workflow is simulator-only; the authors say so clearly and avoid any quantum-advantage claims, so it is fine as an interface test, but it says nothing about hardware feasibility. Minor point: optimal MAXSCI is molecule-dependent and requires calibration, which they acknowledge.\n\nNo fatal flaws. The central argument—the implementation does what it says for the tested inputs—holds up, provided the scope is narrowed to semiquartic force fields. I would send this to peer review with a referee who knows VCI and who insists on an all-distinct quartic test and, if feasible, at least one comparison with experimental fundamentals. For people working on vibrational structure software or quantum-classical spectroscopy workflows, this is worth a look.","headline":"A solid, reproducible VSCF/VCI software paper whose one real novelty—the state-list interface for externally sampled configurations—is worth taking seriously; the main validation gap is that every benchmark uses semiquartic force fields, so the full-quartic coupling path goes unexercised.","tokens_in":24895,"tokens_out":3241,"would_cite":true,"duration_ms":34920,"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":"ViBra computes anharmonic molecular spectra by selected and symmetry-adapted configuration interaction, matching reference water energies to 0.001 cm-1.","keywords":["anharmonic vibrational spectroscopy","vibrational configuration interaction","VSCF","selected CI","Epstein-Nesbet perturbation theory","symmetry-adapted VCI","quantum sampling","quartic force field"],"falsifier":"Run ViBra on a well-characterized small molecule (e.g., water or formaldehyde) using a high-level quartic force field, and compare the predicted fundamental, overtone, and combination band positions and intensities with high-resolution gas-phase experimental spectra; if the mean absolute error exceeds roughly 10-20 cm-1 or a Fermi-resonant intensity pattern is misassigned, the claim of spectroscopically reliable VCI would be contradicted.","tokens_in":23918,"feed_emoji":"⚛️","tokens_out":7235,"duration_ms":63616,"temperature":0.7,"pith_summary":"The paper introduces ViBra, an open-source program that predicts anharmonic vibrational spectra of molecules starting from a quartic expansion of the potential energy surface. It combines a vibrational self-consistent field (VSCF) mean-field step with a subsequent configuration interaction (VCI) step, and offers two ways to keep the VCI matrix manageable: selected VCI, which uses Epstein-Nesbet perturbation theory to keep only the configurations with the largest second-order coupling, and symmetry-adapted VCI, which block-diagonalizes the Hamiltonian using the molecule's point group. The program is validated against independent reference calculations on water and a set of small molecules, with VSCF and full VCI matching reference energies to about 0.001-0.01 cm-1, symmetry-adapted VCI reproducing full VCI to numerical precision, and selected VCI converging variationally to within a few cm-1 of full VCI at reduced cost. As a proof of concept, the paper shows how an externally supplied list of vibrational configurations—here generated by a simulated quantum sampling algorithm—can be fed into ViBra and either diagonalized directly or enlarged through the same perturbation-theory screen before variational rediagonalization. The intended payoff is a practical, freely available bridge between quantum-sampling algorithms and accurate vibrational spectroscopy.","feed_headline":"ViBra matches benchmark water spectra to 0.001 cm-1","feed_subtitle":"Selected configuration interaction keeps anharmonic vibrational calculations fast—and opens a path for quantum-computed seeds.","key_machinery":"The central mechanism is the Epstein-Nesbet second-order perturbative screen (EN-PT2) that drives configuration selection: for each reference state of interest, every external configuration is ranked by the squared magnitude of its Hamiltonian coupling to the reference wavefunction divided by the energy difference, using the exact diagonal Hamiltonian element in the denominator rather than a harmonic estimate. Keeping the top N selections per state yields a compact active space, and the subsequent variational rediagonalization is exact in that space. A second mechanism is symmetry adaptation: for the eight Abelian point groups with one-dimensional irreps, the character of each normal mode is","core_discovery":"On its own terms, the paper's discovery is a validated implementation that makes full vibrational configuration interaction (VCI) practical for molecules with dozens of modes by combining two reductions: an Epstein-Nesbet second-order perturbative screen that selects the most important configurations (S-VCI), and an exact block diagonalization for molecules with Abelian symmetry (SA-VCI). The VSCF and full VCI modules reproduce independent reference water results to within 0.001-0.01 cm-1, SA-VCI matches full VCI to numerical precision across five point groups, and S-VCI reaches variational convergence within about 1-5 cm-1 of full VCI while using a fraction of the Hamiltonian dimension. The","pith_inferences":["The load-bearing comparison in the paper is internal (ViBra vs. established reference implementations and vs. its own full VCI); a systematic quantitative test against measured gas-phase anharmonic frequencies and intensities for the validation molecules would be a natural next step that the current work leaves open.","The same EN-PT2 screen that repairs the quantum seed could in principle also repair a cheap classical guess (harmonic or low-order perturbation theory) with similar effect, which suggests the quantum sampler's role may be to provide a diverse seed in spectrally congested regions rather than a complete one.","Because the quartic force field is the only representation of the potential, the architecture would need higher-order force constants for weakly bound or strongly anharmonic molecules; the paper notes the code is general in this respect but does not test it.","The quantum proof-of-concept uses simulated samples only, so a hardware experiment that measures the shot count and noise levels needed to maintain the reported ~1 cm-1 accuracy after EN-PT2 enlargement would be the decisive test of the hybrid workflow's practicality."],"forward_implications":["Routine anharmonic spectra for molecules with tens of modes: S-VCI brings the Hamiltonian matrix down to tens of percent of the full VCI dimension while keeping low-lying state energies within a few cm-1.","Symmetry-adapted VCI gives exact full-VCI energies at reduced cost for molecules in the supported Abelian point groups; for ethylene the observed wall-clock reduction was roughly 20-fold.","Overtones and combination bands acquire physically correct intensities because the dipole surface is taken to second order, which also redistributes oscillator strength through Fermi resonances.","A quantum sampling algorithm can be plugged into the workflow without changing ViBra: its bitstrings are decoded into vibrational configurations, and the classical EN-PT2 screen repairs the omission errors that otherwise cause spurious state mixing and energy deviations up to 70-80 cm-1 in the raw seed.","The same state-list interface means any external configuration generator, not only quantum samplers, can be used as a seed, making the method a general subspace-projection plus refinement scheme for vibrational structure."],"fun_headline_variants":["ViBra: Full VCI now practical via screening and symmetry","ViBra: Water to 0.001 cm-1, full VCI made practical","ViBra: SA-VCI equals full VCI to numerical precision","ViBra: Perturbative screening tames vibrational CI","ViBra: Quantum-sampled seeds for vibrational CI"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire pipeline inherits the accuracy of the Taylor-truncated quartic (and in practice semiquartic) potential energy surface supplied by the electronic-structure input, and the validation never compares final computed spectra against measured anharmonic frequencies and intensities.","fun_headline_variants_meta":{"raw":{"variants":["ViBra: Full VCI now practical via screening and symmetry","ViBra: Water to 0.001 cm-1, full VCI made practical","ViBra: SA-VCI equals full VCI to numerical precision","ViBra: Perturbative screening tames vibrational CI","ViBra: Quantum-sampled seeds for vibrational CI"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001324,"raw_usage":{"total_tokens":5239,"prompt_tokens":772,"completion_tokens":4467,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":4376}},"tokens_in":516,"tokens_out":4467,"duration_ms":37311,"temperature":1.0,"reasoning_tokens":4376,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T04:22:05.706467+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run ViBra on a well-characterized small molecule (e.g., water or formaldehyde) using a high-level quartic force field, and compare the predicted fundamental, overtone, and combination band positions and intensities with high-resolution gas-phase experimental spectra; if the mean absolute error exceeds roughly 10-20 cm-1 or a Fermi-resonant intensity pattern is misassigned, the claim of spectroscopically reliable VCI would be contradicted.","supporting_citations":[],"review_version":1}