{"id":"6d56022c-3e7b-4d11-b140-a8089396335b","arxiv_id":"2412.01061","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A fractionalized Kohn-Sham scheme using holon and spinon auxiliary particles reproduces DMRG ground states of one-dimensional t-J chains with a simple local density approximation.","lead":"This paper replaces electrons in Kohn-Sham DFT with fractionalized 'holon' and 'spinon' particles to describe strongly correlated electrons, testing the idea on one-dimensional t-J chains. The new scheme matches DMRG ground-state energies and densities for the test cases while running much faster.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The benchmark agreement may be a fit: A and B are tuned on inhomogeneous chains, and the linear ε*xc is defined by that choice, so Figures 3–4 are not an independent test of the KS* LDA.","rationale":"The reader's weakest assumption is essentially correct, and it is the load-bearing point. Standard DFT LDA is parameter-free once the homogeneous-gas data are fixed; here A and B are free parameters tuned on inhomogeneous targets, and the XC functional is a linear fit whose coefficients are not independent of A and B. The fixed A,B applied to two-, three-, and nine-hole cases is encouraging, but those cases are drawn from the same J and potential families used to select A,B, so they are not truly out-of-sample. The claim is plausible and the method is novel, so the appropriate verdict is unchanged: conditional acceptance pending a genuine out-of-sample test and, ideally, released code reproducing the self-consistent loop.","tokens_in":9326,"tokens_out":10612,"duration_ms":100782,"concrete_test":"Run a hold-out validation with the published parameters fixed: use only the homogeneous-chain data and a subset of benchmark potentials (e.g., OBC and impurity chains) to set A, B, and the ε*xc line; then, without any re-fitting, compute GS energies and densities for the held-out periodic-potential chains and for J/t=0.5 and J/t=1.0 systems, comparing to DMRG. If the held-out per-site energy error exceeds 0.01t or the density MAE exceeds 0.01, the reported accuracy is at least partly a consequence of parameter tuning.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central numerical claim depends on the LDA functionals A, B, and the linear ε*xc(n)=−0.21n+0.12 transferring from homogeneous chains to arbitrary inhomogeneous potentials. The paper concedes, after Eq. (11), that 'Testing calculations on inhomogeneous chains show that A=0.45 and B=−0.05 generally perform well' before presenting the benchmark figures. Because ε*xc is defined as ε_GS−t_KS*, and t_KS* itself depends on A and B, the choice of A and B also fixes the XC functional; the parameters are not determined by the homogeneous-chain input alone. The additional two-, three-, and nine-hole tests use the same J/t=0.3 and the same three potential families (OBC, single-site impurity, periodic cosine), so they do not break the circularity. The formal exactness of the KS* auxiliary system is asserted without proof, and the variational argument around Eq. (12) is a stationarity condition, not a demonstration that every density is representable by Eq. (7). Therefore the reported agreement is not yet evidence that the KS* construction, rather than the tuned A, B, and ε*xc fit, is responsible for the DMRG-level accuracy.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a 'fractionalized Kohn-Sham' (KS*) scheme for strongly correlated electrons, using the one-dimensional t-J model as a test bed. The auxiliary system is a noninteracting mixture of holons and spinons with density-dependent hoppings t^h_i(n), t^f_i(n) and an exchange-correlation potential V_xc,i(n). LDA forms are constructed from homogeneous-chain data using two parameters A and B, and a linear fit for epsilon_xc*(n). Self-consistent solutions are then compared against DMRG for one-hole ground-state energies as functions of chain length, impurity strength, and potential period, and for one-, two-, three-, and nine-hole density distributions. The authors report that the KS* iterations reproduce DMRG energies and densities at much lower computational cost.","tokens_in":9537,"tokens_out":6247,"duration_ms":59693,"significance":"The central idea is original and potentially consequential: if a fractionalized auxiliary system can play the role that the noninteracting electron system plays in standard Kohn-Sham theory, it would open a new route to DFT for strongly correlated materials. The multihole calculations without refitting A, B, or the XC fit are genuine evidence of predictive power, and the decomposition of the total energy into T_KS*, E_xc*, and E_ext usefully shows that the XC contribution is essential. The numerical comparisons with DMRG are plausible and cover several physically relevant inhomogeneous setups. However, the parameter selection is partly circular with respect to the benchmarks, the formal existence of the KS* density functions is asserted rather than proved, and no quantitative error estimates are provided. These issues are load-bearing for the claim of DMRG-level accuracy and for the 'formally exact' characterization of the scheme.","major_comments":[{"comment":"The LDA parameters A=0.45 and B=-0.05 are selected by testing on inhomogeneous chains before the benchmarks are presented, and the linear XC functional epsilon_xc*=-0.21 n + 0.12 is fitted to t_KS* computed with those same parameters. Because epsilon_xc* is defined as epsilon_GS - t_KS*, the choice of A and B also fixes the XC functional, so Figures 3 and 4 do not provide an independent test of the transferability of the LDA functionals. The two-, three-, and nine-hole cases use the same J/t=0.3 and the same three potential families, so they do not break this circularity. I request an explicit out-of-sample validation: for example, hold out one potential family or a different J/t during parameter selection and report the errors on the held-out cases; and state how many and which inhomogeneous chains were used in the tuning.","section":"After Eq. (11), 'The values of A and B are determined...' and Figs. 3-4"},{"comment":"The formal claim that 'function forms ... in principle exist' such that Eq. (7) reproduces the exact ground-state density is asserted without proof. The constrained-search references [21,22] apply to the original electron Hamiltonian, not to the fractionalized auxiliary system in Eq. (7). The variational argument in Eq. (12) is a stationarity condition for a fixed auxiliary Hamiltonian; it does not establish that every density of the interacting t-J chain is representable by Eq. (7) with local density functions t^h_i(n), t^f_i(n), and V_xc,i(n), nor does it show that the corresponding energy functional has a variational minimum that equals the exact ground-state energy. The authors should either provide a proof or clearly restate this existence as an assumption that the numerical results are intended to test rather than derive.","section":"Eq. (7) and the KS* statement; Eq. (12)"},{"comment":"The central numerical claim is not supported by any quantitative error metric. The manuscript reports curves and density distributions but never gives the maximum or mean absolute deviation of E_GS or n_i from DMRG, nor the DMRG bond dimensions or truncation errors at each benchmark point. I request explicit numerical errors for every benchmark case, together with DMRG convergence parameters, so that 'comparable to DMRG' can be assessed quantitatively rather than visually.","section":"Figs. 3-4 and the statement 'accuracy comparable to DMRG'"}],"minor_comments":[{"comment":"There is a typo in 'the general concept the of KS * scheme'; it should read 'the general concept of the KS* scheme'.","section":"Introduction, after Eq. (2)"},{"comment":"The text says 'as defined by the deference between epsilon_GS and t_KS*'; 'deference' should be 'difference'.","section":"Section on A and B determination"},{"comment":"Figure 1(c) would be much easier to read with a legend and axis labels identifying t_h, t_f, V_xc^T, and V_xc^I; the current caption does not explain the symbols used for each quantity.","section":"Fig. 1 caption and panel (c)"},{"comment":"The sentence 'the KS* formalism is able to well trace the DMRG GS energy' is ungrammatical; consider 'the KS* formalism reproduces the DMRG ground-state energy well'.","section":"Section on self-consistent iteration"},{"comment":"The Supplemental Material is referenced for implementation details and additional benchmarks, but it is not included in the arXiv submission; the implementation in Sec. IV and the additional comparisons in Sec. V are therefore not verifiable from the preprint alone.","section":"Supplemental Material reference"}],"recommendation":"major_revision","confidential_remarks":"The paper is suitable in scope for a physics letter, but the tuning of A and B on the same inhomogeneous chains used for benchmarking is the main correctness risk. I would not reject, because the multihole results without refitting are encouraging and the missing formal proof can be reframed as an assumption. The revision should add out-of-sample tests and quantitative errors, and should soften the 'formally exact' wording accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a genuinely new angle on DFT for strongly correlated systems, and the 1D demonstrations are clean enough to justify a serious look. But the paper's headline claim depends on parameters chosen against the benchmark class, so it should be read as a proof-of-principle, not a validated transferable approximation.\n\nWhat's new: replacing the noninteracting electron auxiliary system with a holon-spinon system whose hoppings and XC potential are density functionals. That's not in the cited literature. The decomposition into T_KS* and Exc*, and the demonstration that Exc* matters (the mean-field gets it wrong), is a nice piece of insight. The DMRG comparisons for one-, two-, three-, and nine-hole chains are carefully done, and the claimed speedup (an hour on a cluster vs a minute on a laptop) is plausible for a 1D lattice problem solved by diagonalizing a noninteracting system. The paper is honest about the tuning: it states that A=0.45, B=-0.05 'generally perform well' on inhomogeneous chains.\n\nSoft spots. The formal status of the KS* functionals is asserted, not proved. The variational argument around Eq. (12) is a stationarity condition; it doesn't show that every density, or even the relevant densities, is representable by Eq. (7). That's not fatal for an LDA-style paper, but it means the 'in principle exact' claim is a postulate. More importantly, the benchmark agreement is partly tuned: A and B are chosen on the same class of inhomogeneous chains, and epsilon_xc* is defined through t_KS*, which already depends on A and B. So Figures 3 and 4 are not an independent test of the LDA. The multihole cases use the same J/t and the same three potential families, so they don't fully break the circularity. The paper would be stronger with a test on a different J/t or a different potential shape, or with quantitative error bars. None of this kills the idea; it just means the current evidence supports 'promising construction' rather than 'validated approximation'.\n\nBottom line: this deserves peer review. It's a fresh idea, the numerics are real, and the limitations are acknowledged. A referee should push for a cleaner out-of-sample test and a more careful statement about exactness, but the core proposal is worth engaging with. I'd bring it to reading group.","headline":"A fresh and promising fractionalized-KS construction with honest 1D tests, but the benchmark agreement is partly tuned and the exactness claim is asserted; send to review and push for out-of-sample tests.","tokens_in":10122,"tokens_out":2006,"would_cite":true,"duration_ms":19103,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.15.Mb","71.27.+a"],"model":"deepseek-v4-flash","headline":"The paper claims that density functional theory can be extended to strongly correlated electrons by using a noninteracting auxiliary system of fractionalized particles—holons and spinons—instead of electrons, and demonstrates the idea on…","keywords":["density functional theory","Kohn-Sham scheme","fractionalized quasiparticles","spin-charge separation","t-J model","strongly correlated electrons","local density approximation","DMRG benchmark"],"falsifier":"Solve the KS* equations for an inhomogeneous t-J chain outside the tested family—for instance a 100-site chain with a steep double-well potential, a random disordered potential, or a large periodic superlattice—and compare the predicted density and total energy with DMRG calculations on the same chain. If the deviations grow beyond DMRG accuracy as the inhomogeneity strengthens, the transferability of the homogeneous-chain LDA parameters is the point of failure.","tokens_in":9029,"feed_emoji":"⚛️","tokens_out":7676,"duration_ms":61321,"temperature":0.7,"pith_summary":"The paper proposes a new Kohn-Sham scheme, called KS*, in which the auxiliary noninteracting system is composed of fractionalized particles—spinless holons carrying charge and neutral spin-1/2 spinons—rather than ordinary electrons. The rationale is that strong correlations, such as those in doped Mott insulators, obscure the electron reference but leave weakly interacting quasiparticles. Using the one-dimensional t-J model as a test bed, the paper shows that a simple local density approximation built from homogeneous-chain data reproduces DMRG ground-state energies and densities for inhomogeneous chains, at a far lower computational cost. If correct, this opens a path for density functional calculations in the strong-correlation regime where conventional Kohn-Sham schemes are known to fail.","feed_headline":"Holons and spinons bring DFT to the strong-correlation regime","feed_subtitle":"A Kohn-Sham scheme built on fractionalized particles matches DMRG on t-J chains at far lower cost.","key_machinery":"The central object is the fractionalized auxiliary Hamiltonian $\\hat H_{KS*}$ (Eq. 7), in which the electron is replaced by the slave-particle decomposition $c_{i\\sigma}=h^\\dagger_i f_{i\\sigma}$: spinless hard-core boson holons carry the charge and neutral spin-1/2 fermionic spinons carry the spin. The paper's innovation is to make the holon hopping $t^h_i$ and spinon hopping $t^f_i$ density-dependent, with LDA forms derived from the homogeneous-chain correlations, and to augment the usual interaction exchange-correlation potential $V^I_{xc}$ with a kinetic-generated potential $V^T_{xc}$ that comes from differentiating the density-dependent kinetic operator. Standard Kohn-Sham logic is then applied: solve the auxiliary system self-consistently, enforce the single-occupancy constraint with $\\lambda_i$, and obtain the interacting ground-state energy as $T_{KS*}+E_{xc*}+E_{ext}$. The matching to DMRG for one- and multi-hole inhomogeneous chains is the evidence that this auxiliary representation captures the physics.","core_discovery":"The paper's central claim is that a Kohn-Sham-like density functional scheme can be based on a fractionalized auxiliary system of holons and spinons and that, with a local density approximation, it delivers ground-state energies and densities for the inhomogeneous 1D t-J model that agree with DMRG. The auxiliary Hamiltonian (Eq. 7) uses density-dependent holon and spinon hoppings $t^h_i$ and $t^f_i$, an exchange-correlation potential $V_{xc}$, and local chemical potentials $\\lambda_i$ that enforce the no-double-occupancy constraint as one fractionalized particle per site. The paper determines the holon and spinon correlation functions by LDA forms $\\langle h^\\dagger_i h_{i+1}\\rangle \\approx A(1-n_i)$ and $\\langle f^\\dagger_{i\\sigma} f_{i+1\\sigma}\\rangle \\approx 1/\\pi + B(1-n_i)^2$ with $A=0.45$, $B=-0.05$, and takes the homogeneous-chain XC energy per electron as a linear fit $\\epsilon_{xc*} = -0.21 n + 0.12$. The reported results include accurate one-hole energies across chain lengths, impurity strengths, and periodic potentials, multi-hole density profiles for two, three, and nine holes, and a runtime advantage: DMRG takes over an hour on a 24-core node for a 50-site chain while the KS* iteration finishes within a minute on a laptop.","pith_inferences":["If the KS* construction transfers to other models, it would imply that the failure of conventional DFT for Mott systems is not the density functional idea itself but the choice of noninteracting electron reference; any weakly interacting fractionalized reference with a good LDA could work.","A natural testable extension is to apply KS* to Hubbard ladders or the two-dimensional t-J model at moderate sizes, where DMRG is still reliable enough to provide a benchmark and the emergent gauge field becomes necessary.","The density-dependent hoppings generate an extra XC contribution via functional differentiation; the same term should appear in any DFT formalism whose kinetic energy operator depends on density, offering a diagnostic for such schemes."],"forward_implications":["For 1D t-J chains, KS* with simple LDA matches DMRG ground-state energies and densities while reducing the computational cost from over an hour on a 24-core node to under a minute on a laptop.","The XC energy contributes substantially to the total energy in KS*, which demonstrates that slave-particle mean-field treatments miss an essential piece that a density functional can supply.","The success motivates carrying the fractionalized auxiliary-system idea into higher dimensions, where an emergent gauge field must be incorporated, for example by combining KS* with composite-fermion DFT techniques.","In the presence of competing phases, the paper proposes a family of KS* equations, each reproducing the ground-state energy of a specific phase, with the lowest-energy solution selecting the stable phase."],"supporting_citations":[{"why":"Defines the standard Kohn-Sham scheme this paper reformulates for fractionalized particles.","marker":"[3]"},{"why":"Supplies the slave-particle construction $c_{i\\sigma}=h^\\dagger_i f_{i\\sigma}$ and the t-J model background for doped Mott insulators.","marker":"[17]"},{"why":"Establishes DMRG as the near-exact benchmark for 1D systems that the KS* results are compared against.","marker":"[20]"},{"why":"Provides the density-matrix renormalization group algorithm used for the benchmark data.","marker":"[28]"},{"why":"Gives the Luttinger liquid theory that underlies the spin-charge separation assumption.","marker":"[37]"},{"why":"Provides the mean-field slave-particle Hamiltonian from which the LDA forms of the density-dependent hoppings are motivated.","marker":"[38]"},{"why":"Demonstrates the related composite-fermion DFT for fractional quantum Hall systems, cited as the path for incorporating gauge fields.","marker":"[43]"}],"fun_headline_variants":["KS*: DFT for strong correlations via holons and spinons","Fractionalized DFT matches DMRG on t-J chains at laptop speed","Holons and spinons make DFT accurate for correlated electrons","DFT with fractional particles: DMRG accuracy for t-J chains","Spinons and holons: a new DFT for correlated electrons"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The scheme assumes that a noninteracting auxiliary system of holons and spinons with density-dependent hoppings and a local exchange-correlation potential exists and reproduces the exact density of the interacting t-J chain, and that the LDA parameters fitted on homogeneous chains transfer to arbitrary inhomogeneous potentials.","fun_headline_variants_meta":{"raw":{"variants":["KS*: DFT for strong correlations via holons and spinons","Fractionalized DFT matches DMRG on t-J chains at laptop speed","Holons and spinons make DFT accurate for correlated electrons","DFT with fractional particles: DMRG accuracy for t-J chains","Spinons and holons: a new DFT for correlated electrons"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00072,"raw_usage":{"total_tokens":3224,"prompt_tokens":929,"completion_tokens":2295,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":545,"completion_tokens_details":{"reasoning_tokens":2204}},"tokens_in":545,"tokens_out":2295,"duration_ms":16144,"temperature":1.0,"reasoning_tokens":2204,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T04:43:24.878921+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve the KS* equations for an inhomogeneous t-J chain outside the tested family—for instance a 100-site chain with a steep double-well potential, a random disordered potential, or a large periodic superlattice—and compare the predicted density and total energy with DMRG calculations on the same chain. If the deviations grow beyond DMRG accuracy as the inhomogeneity strengthens, the transferability of the homogeneous-chain LDA parameters is the point of failure.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Luttinger liquid theory that underlies the spin-charge separation assumption."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the mean-field slave-particle Hamiltonian from which the LDA forms of the density-dependent hoppings are motivated."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates the related composite-fermion DFT for fractional quantum Hall systems, cited as the path for incorporating gauge fields."}],"review_version":1}