{"id":"f5139b9e-9716-4ee1-8273-dbb5a2909b43","arxiv_id":"1908.03068","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A proof-of-principle that inverse Kohn-Sham inversion can extract local potentials from nuclear densities, with reliable interior results but unreliable tails for experimental sum-of-Gaussian densities.","lead":"This paper tests two algorithms that reconstruct the effective potential of a nucleus from its measured proton and neutron densities, the nuclear version of a known electronic-structure problem. It finds the methods reproduce model densities well and give plausible potentials in the nuclear interior, but that the Gaussian tails used to fit experimental densities make the potential unreliable at large distances.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Tail artifact may contaminate the claimed reliable interior potential; the paper never tests whether the SoG Gaussian tail biases the interior, since vLB/CV agreement is not independent validation.","rationale":"The reader's weakest assumption is the local/central ansatz and the neglect of spin-orbit and nonlocal effects. I regard that as less load-bearing than the tail-contamination issue, because the exact Kohn-Sham potential is by construction local and multiplicative; the benchmark already demonstrates that a purely local central potential can reproduce an HF density generated with nonlocal effective mass and spin-orbit terms. The less guarded point is the paper's positive claim that the interior potential is reliable. The inversion is ill-posed, and the paper itself flags tail pathologies; yet no test is performed to show that the unphysical Gaussian tail is confined to the tail region. Since both inversion algorithms are fed the same target density, their agreement cannot certify spatial locality of the artifact. A hybrid-density numerical experiment would settle this directly. The existing CONDITIONAL verdict is appropriate, and the specific condition of quantifying tail-to-interior contamination should be made explicit; therefore the reader's verdict does not need to change.","tokens_in":15992,"tokens_out":10134,"duration_ms":132044,"concrete_test":"Use the SkX-HF densities from Sec. IV for 40Ca and 208Pb. Construct a hybrid target density that is exactly the HF density for r < R0 (e.g., 8 fm for 208Pb) but whose tail beyond R0 is replaced by a Gaussian of the same form and width as the experimental SoG tail. Run both vLB and CV inversions on this hybrid density and compare the interior and surface potentials (r < R0) with the benchmark HF potential. If the interior potential shifts by more than the ~2.5 MeV benchmark deviation, the SoG tail contaminates the interior and the 'reliable except tail' conclusion fails. A complementary sensitivity check is to perturb the SoG coefficients within the quoted experimental uncertainties and recompute; if the interior potential changes by more than a few MeV, the paper's lack of uncertainty propagation is decisive.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's operative claim is that the extracted potential is reliable in the interior and at the surface, with only the tail unreliable. The supporting evidence is (i) agreement between the vLB and CV methods and (ii) a benchmark against SkX-HF densities. Neither establishes interior reliability for SoG experimental input. vLB and CV are driven by the same target density and the same local-potential ansatz, so their mutual agreement is a consistency check, not an independent validation. The HF benchmark uses physical exponential tails, whereas the experimental input has Gaussian tails that the paper itself identifies as the cause of the divergent oscillator-like potential. The tail can in principle feed back into the interior because the Schrödinger equation is a global boundary-value problem: the divergent potential changes the spectrum, the asymptotic normalization, and hence the occupied orbitals everywhere. Moreover, the vLB update in Eq. (10) divides by the target density, so small SoG parameterization artifacts in the tail are strongly amplified; the authors note the tail divergence but never quantify whether the interior potential has already been biased by the tail or by unpropagated experimental density uncertainties. Thus the central claim, 'reliable information about the potential, except for its tail,' is not yet demonstrated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a first application of the inverse Kohn-Sham (IKS) problem to nuclear systems. Starting from target neutron and proton densities for the spherical doubly-magic nuclei 40Ca and 208Pb, the authors extract a local, central Kohn-Sham potential using two algorithms: the van Leeuwen-Baerends iterative method (vLB) and a constrained variational method (CV). The algorithms are first benchmarked against Hartree-Fock densities from the Skyrme interaction SkX, reproducing target densities to better than 10^-5 fm^-3 and potentials within about 2.5 MeV. The same procedures are then applied to experimental sum-of-Gaussians (SoG) densities for protons in 40Ca and for neutrons and protons in 208Pb. The two methods agree closely with each other and produce potentials that look physical in the nuclear interior, but the potentials diverge in the asymptotic tail because the SoG parameterization has Gaussian tails. The authors conclude that the inversion is robust for experimental SoG densities and gives reliable information about the potential except for its tail, and they discuss perspectives for extending the approach.","tokens_in":16254,"tokens_out":5855,"duration_ms":67526,"significance":"If the central claim is established, this is a useful first step: it introduces a concrete numerical pipeline through which experimental or ab initio densities could constrain nuclear Kohn-Sham potentials and, eventually, energy density functionals. The paper's strengths include the use of two methodologically distinct inversion algorithms, a quantitative benchmark against independent Hartree-Fock densities, the explicit fixing of the potential scale through experimental separation energies, and the transparent acknowledgement of the SoG tail problem and of the limitations of the local-density ansatz. The numerical reproduction of target densities is carefully quantified in Tables I and II. However, the experimental conclusion that the extracted potential is reliable in the interior and at the surface is not yet fully supported, because the tail artifact could in principle feed back into the interior and because the two algorithms do not provide independent validation for the experimental input.","major_comments":[{"comment":"The claim that the experimental SoG input yields reliable potentials 'except for its tail' is not demonstrated. The agreement between vLB and CV is a consistency check, not an independent validation, because both methods use the same target density and the same local-potential ansatz. The benchmark in Sec. IV uses Hartree-Fock densities with physical exponential tails, so it does not exercise the Gaussian-tail regime that the paper identifies as the source of the divergent potential. Since the Kohn-Sham equation is a global boundary-value problem, a divergent tail potential can in principle alter the occupied orbitals and hence the potential in the interior. The paper should quantify this effect, for example by replacing the SoG tail beyond the outermost Gaussian with an exponential tail, or by truncating the target density at various radii, and checking whether the interior potential remains unchanged within the claimed accuracy.","section":"Sec. V (concluding paragraph)"},{"comment":"The central ansatz that spin-orbit and non-local effects can be neglected because they are 'not expected to markedly change the KS orbitals' is asserted rather than tested for the experimental cases. The SkX benchmark has an effective mass close to the bare mass (m*/m between 0.92 and 1.08 for 208Pb), so it does not validate the ansatz for functionals with stronger non-locality or spin-orbit coupling. The paper itself acknowledges in Sec. VI that 'there is no guarantee that a purely local effective potential is the correct choice.' This is a load-bearing limitation: if the exact nuclear Kohn-Sham potential is nonlocal, the local potential that reproduces the density is not the Kohn-Sham potential of the exact system. The abstract, Sec. V, and the concluding remarks should therefore consistently qualify the extracted potential as reliable only within the local, central-potential ansatz; the current wording overstates the physical meaning of the result.","section":"Sec. III (first paragraph) and Sec. VI"},{"comment":"The vLB update in Eq. (10) divides by the target density, so small SoG parameterization artifacts in the low-density tail are strongly amplified. The authors note the resulting tail divergence, but they do not propagate the experimental density uncertainties from Refs. [49] and [50] into the extracted potential, nor do they test the sensitivity of the interior potential to the SoG parameters. Without such an uncertainty or sensitivity estimate, the statement that the interior and surface potentials are 'reliable' is not quantitative. A simple test would be to vary the SoG coefficients or the number of Gaussians within the experimental errors and report the spread of the extracted potential in the interior.","section":"Eq. (10) and Sec. V"}],"minor_comments":[{"comment":"The symbol 'Ul' for the centrifugal potential is easily confused with the Kohn-Sham potential U(r); it should be typeset as U_l or defined more prominently.","section":"Eq. (8)"},{"comment":"The captions refer to 'left figure' and 'right figure'; these should be 'left panel' and 'right panel' for clarity.","section":"Figs. 1, 2, 4, 5"},{"comment":"The statement that spin-orbit energy splittings 'have been checked to have no special influence' is not backed by any displayed result; either show the check or cite a reference for this expectation.","section":"Sec. IV (after Fig. 2)"},{"comment":"The phrase 'We miss a formal proof' should be 'We lack a formal proof' or 'A formal proof is missing'.","section":"Sec. II"},{"comment":"The word 'robust' is vague here; the authors should specify the sense in which the procedure is robust, given that the tail behavior is not physical and the interior reliability is the point to be established.","section":"Sec. V (concluding paragraph)"}],"recommendation":"major_revision","confidential_remarks":"To the editor: This is a solid methods paper and, to my knowledge, the first application of inverse Kohn-Sham inversion to nuclear densities. The main issue is that the central experimental claim—'reliable information about the potential, except for its tail'—needs additional numerical evidence to rule out tail contamination and to quantify uncertainty. This is fixable within the scope of the manuscript, so I recommend major revision rather than rejection. The citation practice appears appropriate and the authors are transparent about the limitations of their local ansatz."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is the first serious attempt to transplant inverse Kohn-Sham inversion into nuclear physics, and it is an honest proof of principle. The benchmark is real, the limitations are stated clearly, and the tail pathology is identified rather than hidden. The one genuinely unaddressed issue is whether the SoG tail artifact contaminates the interior potential; the authors conclude it does not, but the evidence they offer does not settle that.\n\nWhat is actually new: nobody had applied inverse KS (IKS) to nuclear densities before, and that matters for the nuclear EDF program. The authors take two algorithms from the electronic literature - van Leeuwen-Baerends iteration and constrained variational minimization via IPOPT - adapt the update for attractive potentials in Eq. (10), and validate both against SkX-HF densities. The benchmark is the strongest part: densities reproduced to ~1e-5 fm^-3 and potentials within ~2.5 MeV under a stated convergence criterion. That is the right kind of evidence for a proof of principle.\n\nThe experimental application (proton density of 40Ca; proton and neutron densities of 208Pb, from sum-of-Gaussians parameterizations) is handled honestly. The two methods agree with each other, the authors note the potentials oscillate and diverge outside the outermost Gaussian, correlate the divergence with the SoG tail, and explicitly disclaim tail reliability. They also flag the laboratory-vs-intrinsic density subtlety for self-bound systems, which electronic applications never face.\n\nWhere I would push: the 'reliable interior and surface' claim rests on (i) mutual agreement of vLB and CV, and (ii) a benchmark whose tails are physical exponentials. Neither tests whether the divergent tail from the SoG input feeds back into the interior. The Schrodinger equation is a global boundary-value problem, and Eq. (10) divides by the tiny target density in the tail, so amplified tail error can in principle shift orbital normalizations and hence the interior potential. My suspicion is that the effect is small for deeply bound orbitals, but that is a suspicion, not a demonstration. A sensitivity test - truncate the density at the outermost Gaussian, or swap the tail for an exponential one - would settle it. A referee should ask for that.\n\nThe modeling assumptions (local, density-only potential; no spin-orbit; no gradients) are stated up front and are consistent with the 'first step' framing. The benchmark gives partial cover, since the extracted potential tracks the SkX-HF one to within 2.5 MeV. Minor issues: no uncertainty propagation from the experimental densities (the neutron density is model-dependent coming out of proton-scattering analysis), and no code released.\n\nBottom line: this is a solid, honest contribution that deserves publication after revision. Send it out; ask for the tail-sensitivity test and a statement on uncertainties. Workers on nuclear EDF construction and on inverse-problem methods will both get value from it.","headline":"First nuclear inverse-KS paper: honest and well benchmarked, but the 'reliable interior' claim is missing the one sensitivity test that would nail it.","tokens_in":16776,"tokens_out":5652,"would_cite":true,"duration_ms":60440,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["21.60.Jz"],"model":"deepseek-v4-flash","headline":"This paper reports that the nuclear inverse Kohn-Sham problem, extracting potentials from measured densities, works in the interior and surface of 40Ca and 208Pb, with a tail artifact from Gaussian fits.","keywords":["inverse Kohn-Sham problem","nuclear density functional theory","Kohn-Sham potential","sum-of-Gaussians parameterization","40Ca","208Pb","density-to-potential inversion","Hartree-Fock benchmark"],"falsifier":"A decisive experiment would be to invert the proton density of 40Ca obtained from a many-body calculation that includes spin-orbit and non-local terms, and compare the interior potential with the one extracted from the experimental sum-of-Gaussians density; a difference of more than a few MeV would show that the local-density assumption is the weak point, while a persisting large-radius divergence would show the tail artifact is algorithmic rather than a property of Gaussian fits.","tokens_in":15820,"feed_emoji":"⚛️","tokens_out":9355,"duration_ms":88930,"temperature":0.7,"pith_summary":"The paper asks whether the effective Kohn-Sham potential of a nucleus can be recovered from its measured neutron and proton densities, reversing the usual density-functional workflow. It reports that two established inversion algorithms, an iterative update and a constrained kinetic-energy minimization, both reproduce the target densities for the doubly magic nuclei 40Ca and 208Pb. On benchmark densities from a Hartree-Fock calculation with the SkX interaction, the extracted potentials match the reference potential within about 2.5 MeV. On experimental densities parameterized as sums of Gaussians, the two methods agree and yield plausible potentials inside the nucleus, but the Gaussian tails force the potential to diverge like a harmonic oscillator at large radius, so the tail is not trustworthy. The paper thus presents inverse Kohn-Sham inversion as a usable new constraint on nuclear energy functionals, with the caveat that density tails must be better controlled.","feed_headline":"Nuclear densities can be inverted into effective potentials","feed_subtitle":"Test on 40Ca and 208Pb yields reliable interior Kohn-Sham potentials, with a tail artifact from Gaussian fits.","key_machinery":"The central object is the one-dimensional radial Kohn-Sham equation for a local potential $U(r)$, with the density built from the occupied single-particle orbitals, $\\rho(r)=\\frac{1}{4\\pi r^2}\\sum_i n_i u_i^2(r)$. The vLB method iterates the potential through $U^{(k+1)}(r)=U^{(k)}(r)+\\gamma\\left(\\rho^{(k)}(r)-\\tilde{\\rho}(r)\\right)/\\tilde{\\rho}(r)$, adjusting the potential wherever the computed density misses the target. The constrained-variational method minimizes the kinetic energy of orthonormal orbitals under the equality constraint $\\rho=\\tilde{\\rho}$, and the Lagrange multiplier for that constraint is the Kohn-Sham potential. The experimental input is handled with the sum-of-Gaussians parameterization of the densities, and the paper identifies the Gaussian tail of that parameterization as the source of the spurious divergent potential at large $r$.","core_discovery":"On the paper's own terms, the central discovery is that the inverse Kohn-Sham problem, previously applied to electronic systems, can be taken up for atomic nuclei. For 40Ca and 208Pb, the two inversion methods converge to the same potential, and for SkX Hartree-Fock input densities they reproduce the reference potential to within about 2.5 MeV in the interior. For experimental sum-of-Gaussians densities, both methods again agree with each other and give physically reasonable potentials up to the nuclear surface, but beyond the outermost Gaussian the algorithms convert the Gaussian density tail into a quadratic, harmonic-oscillator-like potential that grows without bound. The conclusion is that the inversion is stable and provides reliable information about the potential inside and at the surface of the nucleus, while the tail of the extracted potential is an artifact of the density parameterization and should not be interpreted.","pith_inferences":["The tail artifact is a general warning: the asymptotic form of the density parameterization, not the data alone, controls the extracted potential at large radius; one immediate extension would be to redo the inversion with an exponentially decaying tail and check whether the divergence disappears.","If the local-density assumption is dropped, the correspondence between density and potential is no longer guaranteed; a direct test would be to include a gradient term in the inversion and see whether the interior potential moves by more than the paper's reported few MeV.","The same machinery transfers to other doubly magic nuclei with measured charge densities, so the equal-quality interior potentials obtained for 40Ca and 208Pb could be checked against other phenomenological parameterizations without new formalism."],"forward_implications":["For spherical closed-shell nuclei, measured proton and neutron densities can be used to benchmark local phenomenological potentials in the interior and at the surface.","The absolute scale of the extracted potential is fixed by the experimental separation energy of the last occupied orbital, so the inversion yields energies, not just shapes, once that anchor is used.","Potentials extracted from sum-of-Gaussians densities should not be trusted in the tail; the divergence there is an artifact of the parameterization, not a physical signal.","The next planned step is to apply the inversion to densities from ab initio calculations, which the paper expects to make it possible to extract gradient and spin terms of the energy functional.","Knowing the potential along a continuous path of densities could, in principle, reconstruct the energy density functional itself, and the paper identifies neutron drops as a candidate system for such a path."],"supporting_citations":[{"why":"Supplies the two inversion methods, the vLB and constrained-variational algorithms, and the discussion of well-posedness.","marker":"[19]"},{"why":"Introduces the iterative inversion that becomes the vLB method with the attractive-potential modification.","marker":"[23]"},{"why":"Provides the SkX interaction used to generate the Hartree-Fock test densities and reference potentials.","marker":"[48]"},{"why":"Provides the experimental proton/charge densities of 40Ca and 208Pb in sum-of-Gaussians form.","marker":"[49]"},{"why":"Provides the experimental neutron density of 208Pb extracted from proton scattering.","marker":"[50]"},{"why":"Introduces the sum-of-Gaussians parameterization that is the source of the tail artifact.","marker":"[51]"},{"why":"Establishes the Kohn-Sham independent-particle construction on which the whole inversion is based.","marker":"[18]"},{"why":"Gives the Hohenberg-Kohn density-potential correspondence that justifies trying to recover the potential from the density.","marker":"[5]"}],"fun_headline_variants":["Nuclear inverse KS: densities to potentials, interior reliable, tail artifact","Inverting nuclear densities yields potentials, but watch the tail","First nuclear inverse Kohn-Sham: 40Ca, 208Pb interior potentials recovered","From nuclear densities to potentials: interior works, tail is spurious","Nuclear inverse KS test: interior potentials good, tail artifact from fits"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a purely local, central Kohn-Sham potential depending only on the local neutron and proton densities can faithfully describe nuclei once spin-orbit, non-local, and gradient terms are neglected; if those terms change the orbitals and densities appreciably, the extracted potential is not the potential of an exact Kohn-Sham system.","fun_headline_variants_meta":{"raw":{"variants":["Nuclear inverse KS: densities to potentials, interior reliable, tail artifact","Inverting nuclear densities yields potentials, but watch the tail","First nuclear inverse Kohn-Sham: 40Ca, 208Pb interior potentials recovered","From nuclear densities to potentials: interior works, tail is spurious","Nuclear inverse KS test: interior potentials good, tail artifact from fits"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000515,"raw_usage":{"total_tokens":2499,"prompt_tokens":940,"completion_tokens":1559,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":556,"completion_tokens_details":{"reasoning_tokens":1465}},"tokens_in":556,"tokens_out":1559,"duration_ms":11832,"temperature":1.0,"reasoning_tokens":1465,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:25:12.714644+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive experiment would be to invert the proton density of 40Ca obtained from a many-body calculation that includes spin-orbit and non-local terms, and compare the interior potential with the one extracted from the experimental sum-of-Gaussians density; a difference of more than a few MeV would show that the local-density assumption is the weak point, while a persisting large-radius divergence would show the tail artifact is algorithmic rather than a property of Gaussian fits.","supporting_citations":[{"cited_title":"The two approaches that we discuss below have been origi- nally introduced in Refs","cited_arxiv_id":null,"evidence_quote":"Supplies the two inversion methods, the vLB and constrained-variational algorithms, and the discussion of well-posedness."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the iterative inversion that becomes the vLB method with the attractive-potential modification."},{"cited_title":"Messud, M","cited_arxiv_id":null,"evidence_quote":"Provides the SkX interaction used to generate the Hartree-Fock test densities and reference potentials."},{"cited_title":"(right figure) The Kohn-Sham potentials calculated for neutrons and protons with the inversion methods vLB (red dashed lines) and CV (blue dot-dashed lines) are shown","cited_arxiv_id":null,"evidence_quote":"Provides the experimental proton/charge densities of 40Ca and 208Pb in sum-of-Gaussians form."},{"cited_title":"Titin-Schnaider and P","cited_arxiv_id":null,"evidence_quote":"Provides the experimental neutron density of 208Pb extracted from proton scattering."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the Kohn-Sham independent-particle construction on which the whole inversion is based."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Hohenberg-Kohn density-potential correspondence that justifies trying to recover the potential from the density."}],"review_version":1}