{"id":"8622f024-6b8a-4aac-b076-d2101f246874","arxiv_id":"2509.04832","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A Genesis-to-Starobinsky inflation model in Horndeski gravity shifts the scalar spectral index upward enough to bring Starobinsky inflation into agreement with ACT CMB data.","lead":"The paper builds a cosmological model that combines a Genesis phase with Starobinsky inflation, using modified gravity to avoid a beginning singularity. The model predicts a higher scalar spectral index than standard Starobinsky inflation, which may explain recently observed CMB data from the ACT telescope.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Robustness of the claimed n_s enhancement is not established: it is computed only for the hand-chosen transition functions f(t) and U(t) in Eqs. (26)-(27), with no variation over transition shape or width parameter s.","rationale":"The reader's weakest assumption is precisely the load-bearing point: the transition functions are hand-picked, and the paper's own conclusion concedes that only a particular subclass was explored. I agree with this assessment. The additional observation that the leading-order correction scales as g/s^2 and that s is not scanned strengthens the concern: even within the chosen functional class, the effect size is not robust across the natural uncertainty in the transition width. The absence of a robustness check means the paper supports a conditional statement at best: for this specific genesis-to-inflation matching, the ACT n_s can be accommodated. The central claim of robustness would require either a general proof of sign-definiteness or a scan over transition functions. I do not find a reason to move beyond 'conditional'; no change to the reader's verdict is needed.","tokens_in":17905,"tokens_out":12362,"duration_ms":112445,"concrete_test":"Perform a dedicated robustness scan: replace U(t) in Eq. (27) by the family U_α(t) = 1/[1 + exp(-s(t+α))] (and correspondingly adjust f(t) to preserve the Genesis and Starobinsky asymptotics) and compute Δn_s at N_e = 60, for the same (g,n) grid as in Fig. 4, with all other parameters as in Eq. (51). If the sign of Δn_s is not positive for all α in a range that keeps the transition within the stable region, or if the ACT-compatible window in (g,n) disappears, the 'robust enhancement' claim is disproven.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that Genesis corrections 'robustly enhance' the scalar spectral index rests entirely on the two interpolation functions f(t) and U(t) chosen in Eqs. (26)-(27). No derivation fixes these functional forms, and the numerical scan (Figs. 3, 4, 7) and the analytic effective potential in Appendix B are all built on this single choice. The paper itself concedes in the Conclusion that only 'a particular subclass of transition functions' was studied. This is load-bearing because the sign and magnitude of Δn_s are not guaranteed by the asymptotic limits alone: the leading-order correction in V_ψ (Appendix B) carries an overall g/s^2 prefactor, so the transition width s controls the size of the effect, and the coefficient of the U(t)-step contribution is not sign-definite without computing the full time integrals. A different smooth step with the same asymptotics could plausibly change the sign of the correction or reduce it below the ACT significance; without such a robustness check, the claimed resolution of the ACT tension is not established as a generic property of Genesis-Starobinsky models.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs a non-singular cosmological scenario in Horndeski gravity with three stages: an asymptotically flat Genesis phase, a brief transition, and Starobinsky inflation. The authors propose explicit transition functions f(t) and U(t) to interpolate between the Genesis and Starobinsky Lagrangians, solve the background numerically for one parameter set, and check stability (GS>0, FS>0, subluminal scalar speed). They then scan the parameters n and g, compute the scalar spectral index and tensor-to-scalar ratio using generalized slow-roll formulas, and show that the model can reach the region favored by ACT DR6, whereas Starobinsky inflation alone is disfavored. An analytic leading-order correction is derived in Appendix B and recast as a non-polynomial f(R) correction.","tokens_in":18220,"tokens_out":7283,"duration_ms":71451,"significance":"If the construction is correct and the robustness claims hold, the paper offers a conceptually interesting way to embed Starobinsky inflation in a non-singular Genesis framework and to generate an upward shift in n_s from the pre-inflationary phase. The explicit numerical stability check, the analytic leading-order correction, and the claim that the correction cannot be represented by a finite polynomial in R are valuable and testable elements. However, the central observational conclusion is currently supported only for a particular hand-chosen transition and for a scanned region in (n,g); no likelihood is given, no code or data are released, and the Genesis building block is taken from an unpublished preprint. The paper is therefore more a proof-of-principle than a robust explanation of the ACT data.","major_comments":[{"comment":"The central claim that the Genesis corrections \"robustly\" enhance n_s is supported only for the particular transition functions f(t) and U(t) chosen in Eqs. (26)–(27), and the paper itself concedes in Section 5 that only \"a particular subclass of transition functions\" was studied. Since Appendix B shows that the leading correction carries a g/s^2 prefactor and depends on the full time integrals of f and U, neither the sign nor the magnitude of the shift is guaranteed by the asymptotic limits alone. A different smooth step with the same asymptotics could plausibly reduce or reverse the effect. Please add a robustness study varying the transition shape and the width parameter s, and show that Δn_s>0 and the ACT-compatible region persist, or alternatively temper the claim of robustness.","section":"§3.1, Eqs. (26)–(27)"},{"comment":"Stability in the sense of GS>0, FS>0 and subluminal scalar propagation is demonstrated for the single parameter set in Eq. (30), but the light-blue region in Fig. 7 is selected only by the slow-roll conditions in Eq. (52) and by N_e∈(50,60). It is not shown that ghost and gradient instabilities are absent throughout this entire (g,n) region, nor that the no-go theorem of Refs. [3,4] is avoided there. Please verify stability over the full claimed parameter range, or restrict the allowed region accordingly.","section":"§3.2, Fig. 1, and §4, Fig. 7"},{"comment":"There is an inconsistency in the time-shift used for the analytic leading-order correction. From Eq. (26), f(t)→1 as t→∞, so τ(t)=2f(t)/c+t→t+2/c, whereas Appendix B states τ(t)→t+c/2. Since the terms δA_2^τ, the effective potential Vψ in Eq. (53), and the f(R) expression below it all depend on this shift, the analytic correction as written is not consistent with the full model defined by Eq. (28). Please correct this calculation and recheck the comparison in Fig. 8; until then the analytic leading-order result cannot be used to validate the numerical claim.","section":"Appendix B, Eq. (B.4)"},{"comment":"The paper presents the ACT agreement as an explanatory result, but the observable shift in n_s is obtained by scanning the free parameters n and g, with the other parameters fixed. No likelihood or goodness-of-fit is provided, so the plot in Fig. 7 demonstrates existence of a fitting region rather than a prediction. Please clarify what is genuinely predicted by the model—for example, correlations between n_s, r, and N_e—and what is used to fit the data, or the claimed resolution of the ACT tension will be read as a postdiction.","section":"§4, Figs. 3, 4, and 7"}],"minor_comments":[{"comment":"The mathematical expression \"∑_i c_i R^i\" appears as broken text in the abstract and in the main text; please fix the rendering.","section":"Abstract and Eq. (2)"},{"comment":"The construction of the light-blue allowed region is under-specified: please state explicitly whether it is the union of all scanned points satisfying the slow-roll conditions and N_e∈(50,60), and whether stability has been checked there.","section":"Fig. 7"},{"comment":"The Genesis construction, including its stability and strong-coupling properties, is taken from Ref. [11], which is an unpublished preprint by overlapping authors. Please make the dependence explicit in the text so that the reader can identify which results are established here and which are imported.","section":"Ref. [11]"},{"comment":"The formula for N_max appears to have unbalanced parentheses; please recheck the typesetting and the placement of the closing bracket.","section":"Eq. (41)"},{"comment":"The top axis labels such as \"800c^{-1}\" should be written as \"800 c^{-1}\" or \"800/c\" to avoid ambiguity with a product of c and t.","section":"Fig. 1"}],"recommendation":"major_revision","confidential_remarks":"The paper depends heavily on the unpublished preprint Ref. [11] for the viability of the Genesis stage; before acceptance this dependence should be resolved, either by publication of the companion work or by including the necessary stability and strong-coupling results in the present paper. The topic is appropriate for the journal, and the basic construction is coherent, but the robustness and fitting issues described in the major comments prevent acceptance in the current form."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nShort version: this is a solid, careful construction paper, but the title overclaims. The authors build a stable Genesis-to-Starobinsky transition in Horndeski gravity and show that it can shift n_s upward enough to sit inside the ACT ellipse. That part is real work. What is not established is the word 'robustly' in the abstract: the shift is computed for one hand-chosen pair of transition functions, f(t) and U(t), with no variation over shape or width, and the parameters n and g are scanned to match the ACT data. So the agreement is a demonstration of consistency with data, not a prediction from the model.\n\nWhat is actually new: combining the Genesis phase with Starobinsky inflation in one non-singular evolution, with numerical stability checks (G_S > 0, F_S > 0, subluminal scalar speed) throughout, and a derived leading-order correction to the Starobinsky potential that is genuinely non-polynomial when translated to f(R) (eq. B.6). The appendices are careful and the parameter scan is transparent. The paper also honestly concedes in the conclusion that only a 'particular subclass of transition functions' was studied, which is exactly the weak point.\n\nThe soft spots, in order: (1) The central claim of robust enhancement rests on one transition ansatz. The stress-test note is right: the leading correction carries a g/s^2 prefactor, and the sign and magnitude of the step contribution could change for other smooth steps with the same asymptotics. The authors themselves limit the scope in the conclusion, so the abstract's 'robustly' is stronger than the evidence. (2) The scan over n and g to hit ACT is a fit. The model has enough freedom that the agreement is not surprising. (3) The Genesis building block is the authors' own unpublished preprint [11]. That is not a flaw in itself, but it means the foundation is not yet independently vetted. (4) No code or data file is released, so the numerics can't be checked without reimplementation.\n\nNone of this is fatal. The construction is coherent and the f(R) correction is a useful concrete result for people working on Genesis or on resolving the ACT tension. It just doesn't deserve the 'explain the ACT data' frame as a robust prediction.\n\nVerdict: worth sending to a serious referee. The referees should push on the transition-function dependence and on releasing code, but the paper has enough substance to merit full review.","headline":"A well-built Genesis-to-Starobinsky construction with a genuine non-polynomial f(R) correction, but the 'robustly enhance n_s' claim rests on one hand-picked transition and a parameter scan, so the ACT agreement is a fit, not a prediction.","tokens_in":18669,"tokens_out":2786,"would_cite":false,"duration_ms":24938,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["98.80.Cq","04.50.Kd","98.80.-k"],"model":"deepseek-v4-flash","headline":"A Genesis phase before inflation shifts Starobinsky's predictions into the ACT-preferred region of the n_s-r plane.","keywords":["Genesis cosmology","Starobinsky inflation","Horndeski gravity","scalar spectral index","Atacama Cosmology Telescope","non-singular cosmology","f(R) gravity corrections"],"falsifier":"Recompute $\\Delta n_s=n_s(\\text{Genesis--Starobinsky})-n_s(\\text{Starobinsky})$ at 50--60 e-folds before the end of inflation using a different admissible transition, for instance replacing the logistic $U(t)$ and hyperbolic-log $f(t)$ with error-function or polynomial interpolants that keep $G_S>0$, $F_S>0$, and $u_S\\le1$. If any such transition yields $\\Delta n_s\\le0$, the paper's central robustness claim is false; if the positive sign persists over a family of transitions, the explanation of the ACT data is generic.","tokens_in":17724,"feed_emoji":"🌌","tokens_out":12274,"duration_ms":101990,"temperature":0.7,"pith_summary":"Starobinsky inflation is in tension with the latest ACT-era CMB data: it predicts a scalar spectral index $n_s\\simeq0.965$, while the combination of ACT, Planck, lensing, BAO, and BICEP/Keck data favours $n_s\\simeq0.975\\pm0.003$, disfavouring the vanilla model at about the $2\\sigma$ level. This paper proposes that the mismatch is not a failure of Starobinsky inflation but a signal of what came before it. The authors construct a three-stage Horndeski cosmology that begins in an asymptotically flat Genesis phase, transitions briefly back to general relativity, and then enters Starobinsky inflation. The Genesis phase leaves a calculable correction to the inflationary potential that raises $n_s$ and moves the model inside the observed $n_s$--$r$ contours, while keeping the theory stable and weakly coupled throughout. If correct, the result turns the observational tension into evidence for a non-singular pre-inflationary beginning.","feed_headline":"Genesis before inflation shifts Starobinsky's n_s upward, matching ACT","feed_subtitle":"A stable three-stage Horndeski history leaves a calculable imprint that resolves the 2-sigma mismatch.","key_machinery":"The load-bearing mechanism is the ADM-form Horndeski Lagrangian whose coefficients are $$A_2=\\frac{1}{2}f(t)^{-2\\mu-2-\\delta}\\left(-\\frac{g}{$N^{2}$}+\\frac{g}{$3N^{4}$}\\right)(1-U(t))+\\left(-\\frac{$3M_0^{2}$}{4}+\\frac{1}{$3N^{2}$\\tau(t)^2}+\\frac{$3M_0^{2}$(c\\tau(t))^{2/3}}{$2e^{{2n/3}}$}-\\frac{3(c\\tau(t))^{4/3}$M_0^{2}$}{$4e^{{4n/3}}$}\\right)U(t),\\qquad A_4=-\\frac{1}{2}f(t)^{-2\\mu},$$ with interpolating functions $f(t)=\\frac{c}{2}(\\ln[2\\cosh(st)]/s-t)+1$ and $U(t)=e^{st}/(e^{st}+1)$. Here $f$ acts as the effective Planck mass during Genesis and $U$ suppresses the higher-derivative terms; their asymptotics recover the Genesis Lagrangian as $t\\to-\\infty$ and the Starobinsky Lagrangian as $t\\to\\infty$. The observable shift is then computed from the slow-variation formulas $n_s=1-2\\epsilon_H-\\delta_F-\\eta_s-s$ and $r=16u_S\\epsilon_s$, with the Genesis correction entering through the modified background evolution encoded in the potential above.","core_discovery":"The paper's central claim is that a non-singular Genesis stage preceding Starobinsky inflation leaves a calculable imprint on the inflationary potential, and that this imprint moves the predictions toward the values favoured by ACT. The leading correction to the Einstein-frame potential is $$V_\\psi(\\psi)=\\frac{3}{4}$M_0^{2}$$e^{{-2\\sqrt{2/3}}$\\,\\psi}\\left($e^{{\\sqrt{2/3}}$\\,\\psi}-1\\right)^2+\\frac{$3gM_0^{2}$\\left($e^{{\\sqrt{2/3}}$\\,\\psi}-1\\right)$e^{{-se^n-\\sqrt{3/2}}$\\,\\psi/c-7\\psi/\\sqrt6}\\left(c\\,$e^{{\\sqrt{3/2}}$\\,\\psi}+e^ns\\right)}{$4cs^{2}$},$$ and in $f(R)$ form the same correction contains non-polynomial factors such as $(R/M_0^2+3)^{3/2}$, so it cannot be reproduced by $\\sum_i c_iR^i$ additions. Because the Genesis parameter $g$ is positive, as required for an expanding Genesis phase, the correction always increases the scalar spectral index: for the benchmark parameters $\\mu=0.7$, $\\delta=0.1$, $c=5\\times10^{-5}$, $s=2\\times10^{-5}$, $g=5\\times10^{-8}$, $M_0=10^{-5}$, $n=10.4$, the authors find $\\Delta n_s>0$ for modes that froze out 50--60 e-folds before the end of inflation, shifting $n_s$ from the vanilla Starobinsky value $0.965$ toward the observed $0.975\\pm0.003$ while $r$ stays below the ACT/BICEP bound. The same parameter choices keep the theory free of ghosts and gradient instabilities, with subluminal scalar and luminal tensor speeds throughout the entire evolution.","pith_inferences":["A natural first test is to repeat the numerical pipeline with other smooth interpolations sharing the same asymptotics; the robustness claim stands only if $\\Delta n_s$ stays positive over that family.","The proportionality of the correction to $g$ suggests a broader principle: the expansion rate and duration of the pre-inflationary phase control the spectral tilt. A useful cross-check would be a similar matching onto a different inflationary potential to see whether the tilt shift is a general feature of Genesis initial conditions.","The non-polynomial $f(R)$ correction predicts distinctive higher-curvature behaviour at the transition scale; measuring the running of $n_s$ or the shape of non-Gaussianity in future large-scale CMB surveys could distinguish this scenario from polynomial $R^n$ modifications.","The authors restrict attention to mode freeze-out well after the transition. If a pivot mode instead crossed the horizon during the transition, the correction would be much larger and possibly scale-dependent; computing that case would give a sharp observational test."],"forward_implications":["A positive $\\Delta n_s$ across the allowed $(g,n)$ parameter region explains the ACT-era preference for $n_s\\approx0.975$ without adding fields or tuning the Starobinsky potential.","The non-polynomial form of the correction means bounds on polynomial $R^i$ extensions of Starobinsky inflation do not apply; the scenario must be constrained on its own predictions.","The same construction provides a stable, weakly coupled, non-singular cosmology whose late-time behaviour is Starobinsky inflation, so precision CMB measurements constrain the initial-condition phase rather than only the attractor.","Large values of the transition parameter $n$ continuously recover vanilla Starobinsky inflation, so future data can constrain $n$ and thereby the duration of the pre-inflationary Genesis phase.","The sign of the shift is tied to the sign of $g/c$; a negative $g/c$ would move $n_s$ downward, but such values are excluded because the Hubble parameter must stay positive during Genesis."],"supporting_citations":[{"why":"Supplies the underlying Horndeski Genesis model, its stable parameter space, and the strong-coupling bound used to keep the theory weakly coupled.","marker":"[11]"},{"why":"The no-go theorem that any non-singular Horndeski cosmology must avoid; the stability conditions here are chosen to circumvent it.","marker":"[3]"},{"why":"ACT DR6 power-spectrum data that define the observational window the model is trying to fit.","marker":"[13]"},{"why":"ACT DR6 extended-model constraints that provide the $n_s$--$r$ contours used in the comparison and the 2-sigma tension with vanilla Starobinsky inflation.","marker":"[14]"},{"why":"The original Starobinsky $R+R^2$ model whose predictions are the baseline being corrected.","marker":"[15]"},{"why":"Supplies the slow-variation formulas $n_s=1-2\\epsilon_H-\\delta_F-\\eta_s-s$ and $r=16u_S\\epsilon_s$ used to compute observables in modified gravity.","marker":"[56]"},{"why":"Planck 2018 constraints on inflation combined with lensing and BAO data form part of the P-ACT-LB-BK18 likelihood.","marker":"[35]"},{"why":"BICEP/Keck BK18 B-mode limits are the data that bound the tensor-to-scalar ratio $r$ in the combined likelihood.","marker":"[36]"}],"fun_headline_variants":["Genesis stage nudges Starobinsky's n_s to ACT's sweet spot","Pre-inflation Genesis tweaks Starobinsky to fit ACT","Starobinsky plus Genesis: n_s matches ACT data","Genesis then inflation: ACT-friendly n_s shift","Horndeski Genesis upgrades Starobinsky for ACT"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim of robustness rests on the two hand-picked transition functions $f(t)$ and $U(t)$; the paper verifies the $n_s$ enhancement for this particular matching, and if other viable transition shapes produce a smaller or opposite shift, the ACT explanation is not generic.","fun_headline_variants_meta":{"raw":{"variants":["Genesis stage nudges Starobinsky's n_s to ACT's sweet spot","Pre-inflation Genesis tweaks Starobinsky to fit ACT","Starobinsky plus Genesis: n_s matches ACT data","Genesis then inflation: ACT-friendly n_s shift","Horndeski Genesis upgrades Starobinsky for ACT"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000286,"raw_usage":{"total_tokens":1753,"prompt_tokens":1090,"completion_tokens":663,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":706,"completion_tokens_details":{"reasoning_tokens":579}},"tokens_in":706,"tokens_out":663,"duration_ms":5732,"temperature":1.0,"reasoning_tokens":579,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:26:30.132684+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute $\\Delta n_s=n_s(\\text{Genesis--Starobinsky})-n_s(\\text{Starobinsky})$ at 50--60 e-folds before the end of inflation using a different admissible transition, for instance replacing the logistic $U(t)$ and hyperbolic-log $f(t)$ with error-function or polynomial interpolants that keep $G_S>0$, $F_S>0$, and $u_S\\le1$. If any such transition yields $\\Delta n_s\\le0$, the paper's central robustness claim is false; if the positive sign persists over a family of transitions, the explanation of the ACT data is generic.","supporting_citations":[{"cited_title":"Can Horndeski Genesis be Nonpathological?","cited_arxiv_id":"2503.02626","evidence_quote":"Supplies the underlying Horndeski Genesis model, its stable parameter space, and the strong-coupling bound used to keep the theory weakly coupled."}],"review_version":2}