{"id":"2e28916b-15f0-498e-b35e-212eafcf778c","arxiv_id":"2607.18986","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A 16-state collinear energy mapping isolates the four-spin ring-exchange coupling J_ring, and shows the standard four-state J is ring-renormalized by ±2 J_ring S^2 depending on reference.","lead":"This paper introduces a sixteen-state spin-flip scheme that extracts the four-spin ring-exchange coupling directly from density-functional energies, and shows that the usual nearest-neighbor exchange J is contaminated by ring exchange unless the reference spin arrangement is specified. If correct, it gives a cheap standard way to measure an interaction that matters in cuprates and other square-lattice magnets, and corrects reported J values by up to 12%.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The direct 16-state extraction is not reference-independent in the paper's own main test (27.54 vs 34.52 meV, Table 3); the claimed bare J_ring = 29.57 meV is obtained only after fitting three loop amplitudes in Eq. (13), so 'leaving J_ring alone' holds only within the truncated pair-plus-ring model","rationale":"The reader's weakest assumption is exactly the one that the paper's own main test violates. Eq. (4) is algebraically correct for the Hamiltonian of Eq. (1); the problem is that La2CuO4 is not described by Eq. (1) alone, as the paper's own Table 3 shows. The proposed resolution by Eq. (13) is internally plausible and the stripe2 prediction (1.3 µeV) is a good check, but it converts the method from a direct closed-form extraction into a fitted model of loop corrections. The bilinear routes agree with each other to 0.2% and with Eq. (7), which is strong support for the ring-renormalization framework; however, those routes do not by themselves establish that Eq. (4) is reference-independent. Given that the paper is transparent about the reference dependence and provides a testable correction scheme, the conditional verdict remains appropriate. I agree with the reader that this is an addressable gap rather than a refutation, so I recommend no change to the verdict. I would, however, ask the authors to report the additional ℓ=10/strip-bath check above before presenting the method as a standard tool.","tokens_in":21268,"tokens_out":9517,"duration_ms":99231,"concrete_test":"Add the leading ℓ=10 pure-t strip term, with its bath-spin correlator, to Eq. (13) and refit the five reference-bath values of Fig. 4(b). If the inferred bare J_ring shifts by more than ~0.1 meV, the 29.57 meV value is not converged with respect to the loop-tower truncation, and Eq. (4) is not a reference-independent direct extraction for La2CuO4. If it shifts negligibly, the truncation is vindicated and the remaining issue is only the practical requirement of multiple reference baths.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Eq. (4) cancels constants, molecular fields, and all pair terms, leaving J_ring alone for any reference bath. The paper's own data contradict this for the material used to validate the method: Table 3 gives J_ring = 27.54 meV on the Néel bath and 34.52 meV on the FM bath, a ~7 meV spread reproduced in WIEN2k. The paper acknowledges a pure pair-plus-ring Hamiltonian forbids this and attributes it to six/eight-spin loops. But the resolution is not a direct extraction: Eq. (13) introduces J_e^(6), J_zz^(6), and J^(8) as fitted parameters, and the quoted bare value J_ring = 29.57 meV is the solution of that four-parameter fit to the reference baths, not the output of Eq. (4) alone. The 0.2% agreement among the bilinear routes (Eqs. 9-10) is an internal consistency check within the truncated model, not independent verification of Eq. (4). In SrFeO2 the direct sixteen-state value (0.0526 meV) also disagrees with the mapping value (0.0404 meV, Table 5). Hence the headline method is not yet a black-box direct probe of J_ring for strongly coupled magnets; it becomes one only after a model-dependent loop-tower fit is supplied.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a sixteen-state energy-mapping scheme that extracts the four-spin ring-exchange coupling J_ring from collinear spin configurations. The key identity, Eq. (4), is an exact algebraic consequence of the pair-plus-ring Hamiltonian of Eq. (1): a signed sum over the 2^4 collinear arrangements of a plaquette cancels the energy constant, the bath molecular fields, and all pair couplings. The derivation also predicts that the conventional four-state nearest-neighbor coupling is ring-renormalized by ±2J_ring S^2 depending on the reference bath, and that the next-nearest-neighbor coupling should be reference-independent. The method is applied to T-La2CuO4 and SrFeO2. For La2CuO4, three energy routes agree on J_ring ≈ 32.6–32.7 meV, while the direct sixteen-state extraction is reference-dependent (27.54 meV on the Néel bath, 34.52 meV on the ferromagnetic bath); the authors attribute this to six- and eight-spin loop terms and fit a bare J_ring = 29.57 meV. For SrFeO2, J_ring/J2 = 0.006, providing a negative control. The workflow is implemented in the open-source Mag4 package.","tokens_in":21715,"tokens_out":6044,"duration_ms":58625,"significance":"The algebraic machinery is clean and useful. Eq. (4) is a genuine exact identity within the stated model, and the ring-renormalization prediction of Eq. (5) is confirmed to high precision (0.2% agreement and a 6 µeV non-shift for the diagonal). The paper also supplies two valuable practical contributions: explicit gap and moment diagnostics for validating local-spin mappings, and a negative control (SrFeO2) showing that a plaquette is necessary but not sufficient for ring exchange. The open-source Mag4 package and the reproducible workflow strengthen the paper. However, the headline claim that Eq. (4) leaves J_ring alone 'for any reference bath' is not what the paper's own data show for the strongly coupled material that validates the method; the reference dependence is real and is resolved only by introducing a fitted tower of loop couplings. This limits the direct method precisely in the regime where ring exchange matters most, and the presentation should be reframed accordingly.","major_comments":[{"comment":"Eq. (4) is exact for the Hamiltonian of Eq. (1), but the paper's own validation shows that the direct sixteen-state extraction is not reference-independent: VASP gives J_ring = 27.54 meV on a Néel bath and 34.52 meV on a ferromagnetic bath, with WIEN2k reproducing the spread. The paper acknowledges this and resolves it by introducing the loop model of Eq. (13), whose four fitted parameters produce the quoted bare J_ring = 29.57 meV. That bare value is therefore not the output of Eq. (4) alone; it is the solution of a model-dependent deconvolution. Since La2CuO4 is the central validation material, the abstract and introduction overstate the method when they say that the sixteen-state sum cancels the bath and leaves J_ring alone. The authors should qualify this claim as holding only within the pair-plus-ring model and state prominently that in strongly coupled magnets the direct extraction","section":"Sixteen-state extraction: a genuine reference dependence (Table 3)"},{"comment":"The decomposition that recovers the bare J_ring = 29.57 meV is based on a four-parameter fit (J_ring, J_e^(6), J_zz^(6), J^(8)) to five reference-bath values. The paper does not report the residuals of the fit, the parameter uncertainties, or the condition number/collinearity of the correlator matrix. The 1.3 µeV agreement for the stripe2 prediction is a strong consistency check, but it verifies only that no loop family reaches the out-of-plane Cu sites; it does not verify that the three loop families are sufficient, or that further t'-assisted or ten-spin loops are negligible. To support the claim of a converging tower, the authors should provide errors on all four fitted amplitudes and a sensitivity analysis to the choice of included loop families.","section":"Bath decomposition, Eq. (13)"},{"comment":"The text describes 'three independent routes' to J_ring, but Eq. (9) uses the energy-mapping J1 as an input; only Eq. (10) is a pure difference of two four-state calculations. The three routes therefore share data and are not fully independent determinations. The agreement to 0.2% is still a strong internal-consistency check, but it should be described as such rather than as three independent measurements. This matters because the independence of the routes is part of the paper's central validation argument.","section":"Quantitative confirmation of ring renormalization, Eqs. (9)–(10)"},{"comment":"The negative control also shows a discrepancy between the direct sixteen-state value and the mapping value: J_ring = 0.0526 meV from the sixteen-state method versus 0.0404 meV from the energy mapping, and a different supercell gives 0.074 meV with two metallic configurations. This is a 30–80% spread relative to the small mapping value. The paper presents SrFeO2 as a validation of the methodology, but this unexplained scatter in the direct extraction is not discussed. The authors should either explain this reference/cell dependence or explicitly list it as another instance where the direct extraction is not a black-box probe, separate from the La2CuO4 loop decomposition.","section":"SrFeO2 sixteen-state result, Table 5 and footnote d"}],"minor_comments":[{"comment":"When deriving the effective edge coupling ~J_ij, the paper should state explicitly which ring plaquettes contribute to the degree-two term with factor one (the neighboring plaquette) and which contribute to the degree-four term (the chosen plaquette). The connection to the factor 2 in Eq. (5) is then clearer.","section":"Eq. (3) and Supporting Information"},{"comment":"The data points for the five reference baths are shown without error bars. Given that the fitted loop amplitudes are small residuals of near-cancelling energies, error bars or at least a statement of the numerical uncertainty should be included.","section":"Figure 4(b)"},{"comment":"The text notes that the WIEN2k 1×1×2 FM and Néel values differ by 4.7 meV because of different k-meshes, and that a matched-mesh pair 'is being computed.' For a published comparison this is unsatisfactory; either present the matched-mesh result or remove the statement.","section":"Table 6 and WIEN2k comparison"},{"comment":"The manuscript uses J_ring/J1 and J_ring/J2 interchangeably with J_ring/J in places. A consistent notation, plus explicit statement of the shell labeling for the two materials, would avoid confusion.","section":"Notation"},{"comment":"The historical and personal introduction is engaging, but much of it (Whangbo, Hoffmann, 'du bout des doigts') is not needed for the technical content. It could be shortened without loss.","section":"Introduction"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the journal's scope and the algebraic core is sound. The main issue is the overstatement of the direct extraction: the paper's own main result shows that Eq. (4) is not a black-box probe in the strongly coupled material that validates the method, and the bare J_ring is recovered only after a model-dependent loop fit. This is fixable by reframing the claims and by adding robustness information for the fitted amplitudes. I do not see grounds for rejection, but the present version would mislead readers about the range of applicability of the sixteen-state method."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a genuine extension of the four-state mapping to four-spin ring exchange, and the paper’s own numbers show exactly where it works and where it needs help. The algebraic derivation is clean: flipping a plaquette’s four spins through all 16 collinear arrangements and taking the signed sum cancels constants, bath molecular fields, and all pair couplings, leaving J_ring in the pair-plus-ring model. The corollary that ordinary four-state J is ring-renormalized by ±2J_ring S^2 is the kind of sharp, falsifiable claim that makes a paper worth reading. They test it three ways in La2CuO4 — energy mapping, the one-bath deficit, and the two-bath difference — and the three routes agree to 0.2%. The diagonal-coupling prediction, that it should not shift between Néel and FM baths, holds to 6 µeV. That is impressive and not circular: Eqs. (9) and (10) use different energy differences from the mapping. The 16-state scheme on SrFeO2 as a negative control also makes physical sense, and the discussion of why J_ring is tiny there is sensible.\n\nThe soft spots are real but not hidden. The direct 16-state extraction is reference-dependent in La2CuO4: 27.54 meV on the Néel bath, 34.52 meV on the FM bath. The paper openly reports this, attributes it to six- and eight-spin loops, and resolves it by fitting four parameters (bare J_ring plus three loop amplitudes) to five reference baths, with one parameter-free prediction (stripe2 = stripe1) holding to 1.3 µeV. That is a reasonable consistency check, but it is still a fit, not a direct probe. So the abstract’s promise that “J_ring costs no more effort than J” is overstated for strongly coupled magnets; you need multiple baths and a loop-tower model. The SrFeO2 sixteen-state value (0.0526 meV) deviates from the mapping value (0.0404 meV) by about 30%, which is small in absolute terms but makes me cautious about trusting the direct extraction when J_ring is tiny. The Mag4 repository is cited without a commit hash or archived data, which should be fixed before others depend on it.\n\nNone of these are fatal. The paper is transparent about the reference dependence and turns it into a physical diagnostic. The bilinear-route predictions are strong enough that the core framework is very likely right. This is a methods paper for people doing DFT exchange-parameter mapping in cuprates, manganites, and other square-lattice magnets. I’d send it to peer review, with a request that the SI be checked carefully and the code/data be archived properly. I’d probably cite it, though with a note about the black-box claim.","headline":"A real methodological step forward on ring exchange from collinear DFT, with a strong algebraic core and an honest account of where the direct extraction stops being direct; worth a careful referee.","tokens_in":22190,"tokens_out":2477,"would_cite":true,"duration_ms":25689,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.15.-m","75.10.-b"],"model":"deepseek-v4-flash","headline":"A sixteen-state collinear energy sum isolates the four-spin ring exchange from DFT, with a catch when longer loops contribute.","keywords":["four-spin ring exchange","cyclic exchange","energy mapping","sixteen-state method","La2CuO4","SrFeO2","GGA+U","multi-spin interactions"],"falsifier":"Perform the sixteen-state extraction on a well-isolated plaquette in La2CuO4 using five different reference baths; if the measured values do not lie on the plane predicted by Eq. (13) with a single bare J_ring and small loop amplitudes, the model fails. Separately, in any other material with significant ring exchange, compute the four-state nearest-neighbor coupling on both Néel and ferromagnetic references: if their difference is not exactly 4 J_ring S^2 (with J_ring obtained from an independent route such as energy mapping), the central renormalization claim is wrong.","tokens_in":21127,"feed_emoji":"🧲","tokens_out":3592,"duration_ms":77826,"temperature":0.7,"pith_summary":"This paper introduces a sixteen-state energy-mapping scheme that extracts the four-spin ring exchange J_ring directly from collinear GGA+U total energies, much as the standard four-state method extracts bilinear exchange J. The key move is flipping the four spins of a single nearest-neighbor plaquette through all 2^4 arrangements and summing the energies weighted by the product of the four spin signs; every spin-independent constant, bath molecular field, and pair coupling cancels exactly, leaving J_ring alone. As a corollary, the conventional four-state nearest-neighbor coupling is shown to be ring-renormalized by ±2 J_ring S^2 depending on the reference bath, a prediction confirmed in La2CuO4 to 0.2% across three independent routes. The method also reveals its own limitation: when six- or eight-spin loop terms are present, a single reference bath gives an effective, reference-dependent value, which additional baths separate into a bare J_ring plus a converging tower of loop couplings. SrFeO2 provides a negative control where the plaquette exists but J_ring/J is only 0.006, showing that a square plaquette is necessary but far from sufficient for ring exchange.","feed_headline":"Sixteen spin flips isolate the hidden four-spin ring term","feed_subtitle":"A signed sum over one plaquette's collinear states extracts J_ring from DFT—and shows quoted J values can be 12% off.","key_machinery":"The central object is the collinear reduction of the four-spin ring operator: in a collinear state, the bracket in Eq. (1) reduces to J_ring S^4 times the product of the four spin signs. This makes it possible to construct a signed sum over the sixteen spin patterns of a plaquette that cancels all lower-order terms, exactly analogous to the four-state method. The cancellation relies on the fact that two distinct nearest-neighbor plaquettes share at most an edge, so no pair or single-site term can mimic the plaquette product, and the weighted sum over the sixteen states annihilates every lower-degree contribution. The derivation also exposes the ring-renormalized four-state coupling, Eq. (5),","core_discovery":"The central claim is that on any collinear spin configuration the cyclic ring operator collapses to the product of the four spin signs, which allows an exact decomposition of the total energy and a signed sum over the sixteen spin arrangements of a plaquette—Eq. (4)—that cancels all terms below fourth order, leaving J_ring = (1/(16 S^4)) Σ σ_i σ_j σ_k σ_l E_n. The derivation also yields Eq. (5): a four-state extraction of a nearest-neighbor coupling actually returns the ring-renormalized value J_NN − 2 J_ring S^2 on a Néel bath and J_NN + 2 J_ring S^2 on a ferromagnetic bath, so any four-state J quoted without naming its reference is ill-defined in the presence of ring exchange. The paper va","pith_inferences":["A practical consequence the authors leave implicit: for strongly correlated square-lattice magnets, a single-reference sixteen-state J_ring should be treated as an upper or lower bound bracketing the bare value, and at least two or more baths should be used to certify any local quadrilinear probe.","The loop-tower decomposition in Eq. (13) suggests a direct numerical bridge between DFT energy mapping and the t/U expansion of the Hubbard model, where the extracted J_ring and loop amplitudes could be compared with analytical fourth- and sixth-order coefficients.","The orthogonality of the four-spin column to all bilinear columns in the energy-mapping fit explains why J_ring is insensitive to supercell aliasing of pair shells; this robustness could motivate using four-spin ring terms as anchors when fitting complicated magnetic Hamiltonians.","The method could be extended to other plaquette geometries, such as triangular or kagome lattices, where cyclic exchange terms may stabilize non-collinear orders; the same weighted-sum construction, with appropriate collinear spin products, should isolate the corresponding multi-spin couplings."],"forward_implications":["J_ring can now be extracted from standard collinear GGA+U calculations at a cost comparable to the four-state method, since symmetry reduces the sixteen configurations to six or eight inequivalent energies.","Previously reported four-state nearest-neighbor couplings in ring-exchange materials are reference-dependent; in La2CuO4 a four-state J_1 quoted without naming its reference is wrong by 12%.","The difference between ferromagnetic- and Néel-reference four-state couplings provides a cheap diagnostic: it measures, at fourth order, how far a material departs from the pair-plus-ring Hamiltonian.","Using several reference baths resolves the effective J_ring into a bare ring coupling plus contributions from six- and eight-spin loop exchange, giving a hierarchical description of multi-spin interactions.","A square plaquette is a necessary but not sufficient condition for significant ring exchange: SrFeO2 has the same plaquette yet J_ring/J = 0.006, because orbital dilution and spin normalization suppress the cyclic process."],"fun_headline_variants":["16 spin flips reveal hidden four-spin ring exchange","New 16-state method extracts elusive ring term","Ring exchange: why your J might be 12% off","Four-spin ring coupling pinned down with 16 states","Reference matters: ring exchange skews four-state J"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The extraction formula assumes the spin Hamiltonian contains only two-spin exchanges plus the four-spin ring term on the probed plaquette, but the paper's own La2CuO4 test shows that when six- or eight-spin loop terms are present, a single-reference sixteen-state sum returns a reference-dependent effective value rather than the bare J_ring.","fun_headline_variants_meta":{"raw":{"variants":["16 spin flips reveal hidden four-spin ring exchange","New 16-state method extracts elusive ring term","Ring exchange: why your J might be 12% off","Four-spin ring coupling pinned down with 16 states","Reference matters: ring exchange skews four-state J"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000219,"raw_usage":{"total_tokens":1378,"prompt_tokens":943,"completion_tokens":435,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":687,"completion_tokens_details":{"reasoning_tokens":356}},"tokens_in":687,"tokens_out":435,"duration_ms":4583,"temperature":1.0,"reasoning_tokens":356,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T13:47:31.392564+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform the sixteen-state extraction on a well-isolated plaquette in La2CuO4 using five different reference baths; if the measured values do not lie on the plane predicted by Eq. (13) with a single bare J_ring and small loop amplitudes, the model fails. Separately, in any other material with significant ring exchange, compute the four-state nearest-neighbor coupling on both Néel and ferromagnetic references: if their difference is not exactly 4 J_ring S^2 (with J_ring obtained from an independent route such as energy mapping), the central renormalization claim is wrong.","supporting_citations":[],"review_version":1}