{"id":"e4e20c07-753a-495e-8da7-288ccf322d19","arxiv_id":"2501.16628","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A combination of Ricci and inverse-Ricci scalars is constructed that gives regular FLRW background equations, reducing the linear model to general relativity with a shifted Hubble constant.","lead":"This paper proposes a new way to build Ricci-inverse gravity theories by mixing the Ricci tensor with its inverse in a particular combination that avoids, in a symmetric universe, the singular points that normally plague such theories. For the simplest version, the expansion of the universe follows the same equations as general relativity but with a rescaled Hubble rate, while perturbations remain a work in progress.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The singularity-free claim is not established for fluctuations: at de Sitter (w=-1, x=y) the perturbation expansion (21) diverges as 1/(x-y), the tensor action (28) has a 1/φ'^2 divergence, and Eqs. (30)-(31) conflict, so the w∈[-1,1] safety statement in the conclusion is overbroad.","rationale":"The reader's conditional verdict is appropriate. My reading agrees with the identified weakest assumption: the perturbation sector, specifically the de Sitter limit, is unresolved and the conclusion overstates the safe range. The background construction using L_n is algebraically neat and the background EoM for the linear model are indeed GR-like, so the paper advances the program. However, the paper's own equations show that the linearized action has singular denominators at x=y (Eq. 21) and 1/φ'^2 coefficients (Eqs. 24, 28), and the tensor-mode conflict (30)-(31) is acknowledged but left for future work. Since w=-1 lies inside the range for which the conclusion claims safety for fluctuations, the central claim needs a caveat or a full analysis. A conditional acceptance with an explicit exclusion of de Sitter and other φ'=0 points is the fairest outcome; no change to the reader's verdict is needed.","tokens_in":8148,"tokens_out":17272,"duration_ms":164053,"concrete_test":"Take the linear L1 model on an exact de Sitter background (H=const, φ'=0, V=3(1+α)H^2). Compute the tensor-mode quadratic action (28) without assuming the divergent term vanishes; solve the linearized EoM at finite φ' and take φ'→0. If the only finite-action solutions force Eq. (30) and are incompatible with Eq. (31) for k≠0 and α≠0, the theory has no propagating tensor modes at de Sitter and the w∈[-1,1] safety claim fails. A simpler analytic check: substitute a plane wave ansatz into (30) and (31); the two dispersion relations coincide only for v_g^2=1, i.e. α=0, so for any α≠0 the de Sitter limit is singular or strongly coupled.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's strongest claim—freedom from the singularity problem—is demonstrated only at the background level. At the linearized level the theory is not safe over the stated range w∈[-1,1]. Eq. (21) expands L1 to second order with denominators x-y in the perturbation terms (D-2)/(D-1) N0iNi0/(x-y) and -J NijNji/(x-y). At de Sitter (w=-1), one has x=y and φ'=0, so these terms, and the coefficient αJ/φ'^2 in the tensor action (28), diverge unless the perturbation combinations vanish. For tensors, the vanishing requirement is Eq. (30), h''+2Hh'+k^2h=0, while the would-be GR part demands Eq. (31), h''+2Hh'+v_g^2k^2h=0; for α≠0, v_g^2=(1+α/3)/(1+α)≠1, so no nonzero h satisfies both. The paper explicitly acknowledges this conflict and defers it, yet the conclusion asserts safety for fluctuations for -1≤w≤1. Thus the central claim requires either a full treatment of the de Sitter limit or an explicit caveat excluding w=-1 and other φ'=0 points; as written, the assertion that the theory is free from singularities for the relevant cosmic histories is not supported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs a class of Ricci-inverse gravity theories whose Lagrangian is a function of traces of powers of the Ricci tensor, including negative powers (anticurvature). A combined quantity L_n is introduced via an algebraic identity involving R_n, R_{-n}, and R_{-2n}. In FLRW spacetime, L_n simplifies to a closed algebraic form, and for the linear model f=R1+αL1 the background Friedmann equations are shown to coincide with GR up to the rescaling H→√(1+α)H. The paper claims that this construction is free from the singularity problem for both the background and linear perturbations for equations of state with w between -1 and 1. Perturbation actions for scalar and tensor modes are presented, but the scalar constraint is not solved and the de Sitter limit is left with an acknowledged conflict.","tokens_in":8509,"tokens_out":11833,"duration_ms":107157,"significance":"The construction of L_n is elegant, and the background dynamics of the linear model being GR-like is a clean and non-trivial result. The algebraic identity underlying L_n is a useful observation for the Ricci-inverse literature. However, the central safety claim is only demonstrated at the background level. The perturbation analysis, which is essential for the claim, is incomplete: the second-order actions are not derived, the scalar constraint is left unsolved, and the de Sitter tensor sector yields contradictory equations. As presented, the manuscript does not support the conclusion that the theory is free from singularities for fluctuations over the stated range of w. The main value of the paper would be in the background construction if the perturbative issues are either resolved or explicitly excluded from the claims.","major_comments":[{"comment":"The second-order expansion of L1 in Eq. (21) has denominators (x-y). At the de Sitter endpoint (w=-1) one has x=y and φ'=0, so the perturbation terms and the 1/φ'^2 prefactors in Eqs. (24) and (28) diverge unless the associated perturbation combinations vanish identically. The paper's ansatz that these combinations vanish is not proven, and for tensor modes it leads to the contradictory conditions in Eqs. (30)-(31). Therefore the claim that fluctuations are safe for -1≤w≤1 is not supported; the endpoint must either be analyzed or explicitly excluded.","section":"Perturbations, Eq. (21)"},{"comment":"In the de Sitter limit, requiring the singular 1/φ'^2 term in the tensor action (28) to vanish gives Eq. (30), while the remaining GR-like part of the action demands Eq. (31). For α≠0, v_g^2=(1+α/3)/(1+α)≠1, so no nonzero tensor perturbation can satisfy both equations. The paper acknowledges this conflict but does not resolve it; nevertheless the conclusion includes w=-1 in the claimed safe range. This inconsistency is load-bearing for the singularity-free claim and must be resolved before the claim can stand.","section":"Perturbations, Eqs. (30)-(31)"},{"comment":"The scalar constraint equation (25) is not solved; the authors explicitly defer this, noting that solving it would introduce k in the denominator. Since the regularity of scalar perturbations at x=y, φ'=0, and J=0 depends on the behavior of the lapse perturbation A after solving (25), the unsolved constraint leaves the safety claim for scalar modes unverified. The pseudo-GR argument for J→0 does not cover the de Sitter point or generic intermediate values of w.","section":"Perturbations, Eq. (25)"},{"comment":"The second-order actions (24) and (28) are presented as the outcome of 'tedious integration by parts' but no derivation is provided. These actions are the sole basis for the perturbation claims, and they contain the divergent 1/φ'^2 terms. The authors should supply the perturbed Ricci tensor components, the gauge-fixing conditions, and the reduction steps, or at least a detailed outline, so that the presence and form of these singular terms can be independently checked.","section":"Perturbations, Eqs. (24) and (28)"}],"minor_comments":[{"comment":"The algebraic identity (8) and the definition of L_n in (9) are typeset in a garbled way that makes them difficult to read; they should be rewritten with explicit fractions and parentheses.","section":"Specific Model, Eq. (8)"},{"comment":"The variables ξ_N and ξ_NN in Eqs. (11)-(12) are not explicitly defined; please define them as derivatives with respect to ln a.","section":"Specific Model, Eq. (11)"},{"comment":"The definition of J from Eq. (21) should be restated near Eq. (24) for readability.","section":"Perturbations, Eq. (24)"},{"comment":"The phrase 'as long as the EoS parameter w ranges between -1 and 1' should be clarified to state whether the endpoints are included, given that the de Sitter endpoint is exactly where the perturbation analysis encounters the unresolved conflict described in the text.","section":"Conclusion"},{"comment":"There is a typo: 'Talor/Laurent expasion' should be 'Taylor/Laurent expansion'. Additionally, the no-go theorem for L=R1+αR_{-1}^l mentioned in the introduction is stated without a citation; a reference should be provided.","section":"Introduction"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the journal's scope and the background construction is likely to interest the Ricci-inverse community. However, the central claim as stated in the abstract and conclusion goes beyond what the perturbation analysis establishes. If the authors cannot resolve the de Sitter conflict, the claim of singularity-free fluctuations should be narrowed accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Good to have a paper like this to look at. The genuinely new thing is the L_n combination: by mixing R_{-1} and R_{-2} with R_n in the right ratio, the author gets L_n = a^{-2n}(X^n + Y^n). That identity (8) is simple, checkable, and it removes the rational-function singularities that normally plague Ricci-inverse models. The background analysis of the linear model f = R1 + αL1 is also clean: the Friedmann equations reduce to GR with a shifted Hubble parameter Hα = sqrt(1+α) H0. That is a neat result, and I found no error in the background algebra.\n\nThe soft spot is the perturbation sector, and it is exactly where the paper's strongest claim lives. The expansion (21) has denominators x-y and J/(x-y) that diverge at de Sitter (w=-1), where x=y and φ'=0. The tensor action (28) has an αJ/φ'^2 term that diverges there too. The author recognizes this: Eqs. (30) and (31) conflict, and he says detailed analyses are left for future work. But the conclusion still claims the theory is safe for fluctuations for -1 ≤ w ≤ 1. That is overbroad; w=-1 is precisely the de Sitter point. The scalar constraint (25) is also not solved, so the \"pseudo-GR\" reduction rests on an ansatz about J→0 rather than a derivation.\n\nNone of this kills the background construction. The paper is honest about the open issues, and the algebraic core is solid. But the singularity-free claim should be scoped to the background, or the de Sitter/φ'=0 cases need an actual treatment. The design is circular in the sense that L_n was chosen to cancel the singularities—that is worth saying—but it is not a flaw in itself; the construction is the contribution.\n\nFor a reader: this is for people working in Ricci-inverse gravity or modified gravity cosmology. It is a short paper with one useful algebraic trick and a clean background result. I would send it to a referee; the perturbation issues are substantive and need to be engaged, but the paper is not incoherent and the central background result is worth checking. With a revised conclusion and a fuller perturbation treatment, it could be a decent contribution. As is, I would not cite it for the safety claim, but I might cite the L_n construction if I worked on these models.","headline":"A genuinely new algebraic construction and a clean GR-like background result, but the singularity-free claim is not supported at the perturbation level, especially at de Sitter.","tokens_in":8978,"tokens_out":2823,"would_cite":false,"duration_ms":27824,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["83D05","83F05"],"pacs":["04.50.Kd","98.80.-k"],"model":"deepseek-v4-flash","headline":"A new Ricci-inverse gravity with two anticurvature scalars claims to be singularity-free and, in the linear model, to reduce to general relativity with a shifted Hubble parameter.","keywords":["Ricci-inverse gravity","anticurvature scalar","modified gravity","FLRW cosmology","singularity-free cosmology","cosmological perturbations","fourth-order gravity","Hubble parameter rescaling"],"falsifier":"Solve the full fourth-order tensor perturbation equations numerically across a transition into the exponential-expansion phase without assuming $\\phi'^{-2}\\to 0$. If no finite, regular solution exists unless the two conflicting equations (30) and (31) are both satisfied with $v_g=1$, the claim that the theory is safe for fluctuations in that phase is refuted.","tokens_in":7950,"feed_emoji":"🌌","tokens_out":13210,"duration_ms":114704,"temperature":0.7,"pith_summary":"This paper tries to show that a gravity theory built from two anticurvature scalars can avoid the singularities that have plagued Ricci-inverse gravity. The key move is a new scalar $L_1$ constructed so that its FLRW form no longer contains the rational denominators of the anticurvature tensors; all apparent poles become removable. In the linear model, the background Friedmann equations become exactly those of general relativity with the Hubble parameter rescaled by $\\sqrt{1+\\alpha}$, so the early- to late-time evolution is regular. The paper further argues that at the perturbation level the theory is fourth-order with an extra scalar degree of freedom, but in certain limits this degree of freedom can disappear and the modes reduce to a 'pseudo GR' form with a modified group velocity. A sympathetic reader would care because it offers a concrete route to singularity-free cosmology without giving up the geometric, curvature-only description of gravity.","feed_headline":"Two anticurvature scalars yield singularity-free cosmology","feed_subtitle":"In the linear version the background equations match general relativity with a shifted Hubble parameter.","key_machinery":"The load-bearing object is the ratio-symmetric scalar $L_n$ defined in equation (9), built from the Ricci powers $R_n$, $R_{-n}$, and $R_{-2n}$ so that its FLRW value reduces to $a^{-2n}(X^n+Y^n)$ instead of a rational function with poles. In FLRW, $X = a^2 R^0{}_0$ and $Y = a^2 R^i{}_i/(D-1)$; the three apparent poles $X=0$, $Y=0$, and $X=Y$ are all removable. The unified variational formula (4) with intermediate tensors $P^\\mu{}_\\nu$ and $Q^\\mu{}_\\nu$ keeps the equations compact, and the background system uses the modified Friedmann equation together with energy conservation, mirroring the standard $f(R_1)$ treatment.","core_discovery":"The paper's central claim is that the combination $L_n := (R_n - D R_{-n}/R_{-2n})/(D - R_{-n}^2/R_{-2n})$ forms a scalar with the same dimension as $R_n$ whose FLRW value is simply $a^{-2n}(X^n + Y^n)$, with no rational-function poles. For $n=1$ and with $L_1$ added linearly to the Ricci-scalar action, the background equations of motion reduce to $\\rho = 3(1+\\alpha) H^2$ and $p = -(1+\\alpha)(2\\xi+1) H^2$, i.e. general relativity with the replacement $H_0 \\to \\sqrt{1+\\alpha}\\,H_0$. The three singular points $X=0$, $Y=0$, and $X=Y$ are all removable, so the cosmic deceleration/acceleration boundary and the exponential-expansion (de Sitter) limit are not obstructions. On the perturbation side, the claim is that although the action is fourth-order and carries an extra scalar degree of freedom, in the limits $J\\to 0$ and the exponential-expansion phase the extra mode decouples and scalar and tensor spectra reduce to the GR form with a modified group velocity $v_g$; the paper notes an unresolved conflict in the tensor-mode equations in that phase.","pith_inferences":["Beyond the paper, the $J=0$ redshifts $z\\approx 0.67$ and $1.8$ are structure-formation epochs, so the model predicts a change in the propagation speed of curvature perturbations during those epochs; large-scale-structure surveys could in principle look for this signature.","The same algebraic identity that builds $L_1$ can be applied at higher $n$ or to other combinations of anticurvature scalars, making the construction a general template for singularity-free Ricci-inverse theories rather than a single tuned model.","If the tensor-mode conflict in the exponential-expansion phase persists under a full analysis, the theory would suppress primordial tensor perturbations, so an observed primordial $B$-mode signal at CMB scales would rule out this simple version.","Reading $\\rho_{\\rm eff} = -3\\alpha H^2$ as effective cold dark matter implies that background expansion data alone could constrain $\\alpha$, because the model changes the effective matter density without altering the equation of state."],"forward_implications":["The background evolution of the linear $L_1$ model is identical to general relativity with $H_0 \\to \\sqrt{1+\\alpha}\\,H_0$, so standard cosmological solutions carry over with rescaled expansion rates.","The apparent singularities at $X=0$, $Y=0$, and $X=Y$ are removable, so the deceleration-acceleration transition and the exponential-expansion phase are not barriers for the background as long as the equation-of-state parameter lies between $-1$ and $1$.","At the perturbation level the theory is fourth-order with an extra scalar degree of freedom; in the $J\\to 0$ limit, corresponding to $w=-1/3$ or $w=-1/11$ (redshifts $z\\approx 0.67$ or $1.8$), the extra mode decouples and scalar perturbations follow the GR form with a modified group velocity.","Tensor modes reduce to the same pseudo-GR form in the $J\\to 0$ limit, but in the exponential-expansion limit the two natural tensor equations conflict, leaving the fate of tensor fluctuations open."],"supporting_citations":[{"why":"An earlier workable $f(R_{\\pm 1})$ model whose singular point at the acceleration boundary $x=0$ the present construction removes; acts as the baseline for the perturbation comparison.","marker":"[36]"},{"why":"Analyses $f(R_1,R_{-2})$ models whose equations of motion are complicated; motivates the intermediate-variable method used here to keep the EoM compact.","marker":"[37]"},{"why":"Provides the method used in $f(R_1)$ theory to study and cure singularities in the FLRW background, adapted here to $L_1$.","marker":"[40]"},{"why":"A second $f(R_1)$ reference used for the same singularity-curing approach in background evolution.","marker":"[41]"},{"why":"Work that drew attention to non-local operators, motivating the use of negative-index (anticurvature) scalars in gravitational actions.","marker":"[22]"}],"fun_headline_variants":["Two anticurvature scalars banish cosmic singularities","Anticurvature pair yields singularity-free cosmology","New Ricci-inverse theory dodges singularities with twin scalars","Modified gravity with two scalars sidesteps singularities"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The main bet is that the extra wobbling mode that appears in the perturbed equations quietly disappears in the $J\\to 0$ limit and in the exponential-expansion phase, so the simplified 'pseudo GR' behavior is valid; the paper itself finds conflicting equations for gravitational waves in the exponential-expansion limit, so this bet is not settled.","fun_headline_variants_meta":{"raw":{"variants":["Two anticurvature scalars banish cosmic singularities","Anticurvature pair yields singularity-free cosmology","New Ricci-inverse theory dodges singularities with twin scalars","Modified gravity with two scalars sidesteps singularities"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000947,"raw_usage":{"total_tokens":4018,"prompt_tokens":897,"completion_tokens":3121,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":3052}},"tokens_in":513,"tokens_out":3121,"duration_ms":21990,"temperature":1.0,"reasoning_tokens":3052,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T11:49:38.114157+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Solve the full fourth-order tensor perturbation equations numerically across a transition into the exponential-expansion phase without assuming $\\phi'^{-2}\\to 0$. If no finite, regular solution exists unless the two conflicting equations (30) and (31) are both satisfied with $v_g=1$, the claim that the theory is safe for fluctuations in that phase is refuted.","supporting_citations":[{"cited_title":"Circumventing the Ricci-inverse no-go theorem with complexifiable singularities: a novel dark energy model","cited_arxiv_id":"2409.16529","evidence_quote":"An earlier workable $f(R_{\\pm 1})$ model whose singular point at the acceleration boundary $x=0$ the present construction removes; acts as the baseline for the perturbation comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Analyses $f(R_1,R_{-2})$ models whose equations of motion are complicated; motivates the intermediate-variable method used here to keep the EoM compact."},{"cited_title":"The future evolution and finite-time singularities in $F(R)$-gravity unifying the inflation and cosmic acceleration","cited_arxiv_id":"0804.3519","evidence_quote":"Provides the method used in $f(R_1)$ theory to study and cure singularities in the FLRW background, adapted here to $L_1$."}],"review_version":1}