{"id":"d950465c-841b-4f4b-80e8-55692d20a086","arxiv_id":"2611.03429","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"DBI kinetic domination drives a stiff post-inflation phase with w below unity, softening the blue tilt of the primordial gravitational wave spectrum and expanding the parameter region visible to future detectors.","lead":"A string-inspired scalar field theory with DBI kinetic terms can produce a post-inflation stiff phase with an equation of state below the standard value, which flattens the rise of the gravitational wave spectrum. The effect can push predicted signals above the sensitivity curves of planned detectors, making the early kinetic phase testable.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The DBI softening argument is internally sound, but the observable-signal claim rests on a single phenomenological reheating input, rho_e^r (Eq. 71), whose freedom controls the stiff-phase duration and the 835-point viable region.","rationale":"The reader's weakest assumption matches the point I consider most load-bearing: the fixed phenomenological radiation density rho_e^r and the unspecified reheating mechanism. I checked the DBI derivation itself and found no internal inconsistency. Eq. (54) follows from the DBI energy density and pressure in the X >> V limit, Eq. (60)-(61) consistently produce a constant EoS parameter during scalar kinetic domination, and Eq. (65) follows from the horizon re-entry condition. The numerical spectra in Fig. 4 track the analytic expectation, including the w_tot plateaus controlled by kappa and the duration controlled by f0. The frozen-mode caveat in Eq. (63) and b105-b107 is acknowledged and affects modes above k > 0.1 a_e H_e, i.e., frequencies around the cutoff that are far above the bands of LISA, DECIGO, ET, and SKA used for the PLI crossings, so I do not treat it as the dominant concern. The absence of code or data weakens auditability but is not evidence of an error. The observable-signal part of the claim, including the 835 surviving points and the statement that the spectrum would otherwise remain beyond reach in the canonical case, is conditional on rho_e^r and on the benchmark potential. Since the paper explicitly labels rho_e^r phenomenological and states that a full scan is beyond scope, the honest position is that the mechanism is secure while the detection claim is benchmark-dependent. This does not change the reader's CONDITIONAL verdict.","tokens_in":23729,"tokens_out":21736,"duration_ms":149721,"concrete_test":"Re-run the Fig. 5 scan with rho_e^r varied over, say, 1e-16, 1e-17, and 1e-18 mP^4, and for each surviving DBI point also run the canonical model with rho_e^r tuned to give the same N_reh. If the canonical twin also crosses the same PLI, the observational advantage is not DBI-specific; if the DBI surviving region changes by more than an order of magnitude in f0 across the rho_e^r range, the 835-point claim is an artifact of the benchmark reheating input. For one point, check consistency with the end-of-inflation Hubble scale by computing the gravitational-reheating estimate of rho_e^r from the actual H_e of that point.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central analytic step is internally consistent: under kinetic domination Eq. (54) gives w(phi) approximately 1/gamma, and Eq. (65) maps w in (1/3,1) to a GW slope between 0 and 1. I do not find an error in that derivation. The load-bearing uncertainty is in the observable-signal extension. Section II C treats radiation as separately conserved with rho_e^r = 8.1e-17 mP^4, fixed for every {f0,kappa} in the scan. The stiff-phase duration is roughly N_reh ~ ln(rho_phi,e/rho_e^r)/(3w-1), so rho_e^r directly sets which frequencies in Fig. 5 are enhanced and whether a spectrum crosses a PLI. Because the reheating mechanism is unspecified, rho_e^r is a free dial: shifting it changes the canonical comparison and the 835 surviving points. Moreover, if rho_e^r is meant to arise from a concrete mechanism such as gravitational reheating, it should depend on the end-of-inflation Hubble scale, which the DBI parameters themselves modify; holding Eq. (71) fixed across all points could be internally inconsistent for that mechanism. The paper acknowledges this as a modeling choice and states in b177 that a full parameter scan is beyond scope, so this is a conditional-demonstration issue rather than a mathematical error. The softening result itself does not depend on rho_e^r.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper studies the post-inflationary equation of state and gravitational-wave spectrum of a scalar field governed by the DBI action P=(1/f)(1−sqrt(1−2fX))−V, with a runaway potential that ends inflation and then becomes kinetically dominated. It shows analytically that in this regime the EoS parameter is w≃1/γ (Eq. 54), i.e. smaller than the canonical kination value w=1, and that for an exponential warp factor f=f0 exp(κφ/mP) (Eq. 60) the EoS is approximately constant and lies in the stiff range 1/3<w<1 for √6<κ<√12. Combining the frozen-tensor transfer formula (Eq. 63) with the horizon re-entry condition, it derives the slope d log Ω0GW/d log k = n_t + 2(3w_k−1)/(3w_k+1) (Eq. 65), implying a blue tilt between zero and one during the stiff phase. The authors then fix the potential and the radiation density at the end of inflation, solve the background and perturbation equations numerically from inflation to radiation domination, and scan 100×100 combinations of (f0,κ), keeping 835 points that pass CMB, BBN, LVK and reheating constraints and cross at least one future GW sensitivity curve. Appendix A extends the construction to constant-roll DBI backgrounds with an exponentially evolving sound speed.","tokens_in":24187,"tokens_out":11264,"duration_ms":83932,"significance":"The central analytical result is simple, explicit and internally consistent: I checked Eq. (54) from Eqs. (40)–(41) and the slope formula (65) from the re-entry condition, and found no error. The exponential-warp construction is a legitimate way to realize a period of constant w, and the numerical spectra show the claimed softening compared with the canonical case. If the detectability region were robust against the unconstrained reheating input, this would be a valuable concrete target for LISA/ET/DECIGO/LVK and an interesting way to probe non-canonical kinetic structure with GW observations. The paper is also honest about its modeling assumptions, explicitly labelling ρ_e^r as a phenomenological choice and stating that a full parameter scan is beyond scope. The main weakness is that the observable-signal part remains conditional on that single input rather than on a derived reheating mechanism or a robustness scan; the analytic softening result itself does not depend on ρ_e^r.","major_comments":[{"comment":"The observable-signal part of the paper is governed by a single phenomenological input, ρ_e^r in Eq. (71). The duration of the stiff phase is essentially N_reh ~ (1/(3w−1)) ln(ρ_φ,e/ρ_e^r), so Eq. (71) fixes which frequency band in Fig. 5 is enhanced and whether a spectrum crosses any PLI. Since no reheating mechanism is specified and ρ_e^r is held fixed across the entire {f0,κ} scan, the 835-point viable region is a demonstration for one value of this dial rather than a prediction of the model. Moreover, if ρ_e^r were realized by gravitational reheating, its value would depend on the end-of-inflation Hubble scale and hence on the DBI parameters, so holding it fixed over the whole scan could be internally inconsistent for that mechanism. The authors do acknowledge the modeling choice and the lack of a full scan, but the abstract's 'potentially observable signal' and the Conclusions' viability claim still rest on this input. I ask for either a scan over ρ_e^r, a derivation from a concrete reheating model, or a clear statement that Fig. 5 is a benchmark illustration rather than a model prediction.","section":"Sec. II C / Eq. (71); Sec. III C / Fig. 5","authors_acknowledge":"Preliminary tests show this trend is maintained for values of the inflationary parameters different from Eqs. (70)-(71), although a parameter scan over the parameter space of the full model is beyond the scope of the present work.","authors_acknowledge_p":0.61},{"comment":"The quantitative numerical claims—835 surviving points, the roughly ten-order-of-magnitude range in f0, and the PLI-crossing criterion—are not independently reproducible from the text: no code, data, or numerical metadata (integration method, tolerances, initial-condition handling at the end of inflation, or the exact algorithm for deciding that a spectrum crosses a sensitivity curve) are provided. This matters because these numbers are the numerical headline of the paper. I recommend making the scan outputs and analysis code available, or at least providing the numerical settings and a data table for Fig. 5, so that the BBN integral in Eq. (73) and the PLI-crossing test can be audited.","section":"Sec. III C","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The Conclusions' phrase 'the mechanism remains viable over a substantial region of the explored parameter space' should carry the qualifiers 'for the exponential-warp benchmark' and 'for the fixed value of ρ_e^r in Eq. (71)', since the general softening result does not by itself determine the duration of the stiff phase or the detectable frequency band.","section":"Sec. IV","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The left panel of Fig. 5 shows 835 surviving points as overlapping scatter points, which makes the density of the viable region difficult to assess; a contour or density map, or a machine-readable data table, would make the parameter-space claim much clearer.","section":"Fig. 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The adiabaticity condition in Eq. (18) is typeset in a way that makes the terms involving c_s' and c_s'' hard to parse; adding parentheses or a labelled definition of the ratio would improve readability.","section":"Eq. (18)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text defines the reheating time as the moment ρ_φ(N_reh)=ρ_r(N_reh), but it does not explicitly state that no further radiation production is included after that time; this is implied by Eqs. (47) and (50)–(52), and stating it explicitly would sharpen the assumptions behind the BBN condition.","section":"Sec. II C","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision","confidential_remarks":"The analytic part of the paper is sound and, in my view, publishable: the w≃1/γ result and the slope formula are correct and cleanly presented. The point that needs work before acceptance is the conditioning of the numerical/detectability claim on the unconstrained reheating parameter ρ_e^r and on the benchmark exponential warp factor. This is fixable within the manuscript's scope by adding a robustness scan or by demoting Fig. 5 to a clearly labelled benchmark illustration, so I do not see this as a reject. I found no circularity in the argument: the exponential warp factor is a modeling choice, not a hidden fit to the GW signal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Quick take: the core result is solid. Under DBI kinetic domination, Eq. (54) gives w ~ 1/gamma, and Eq. (65) maps that to a GW spectral slope between zero and one. The derivation is internally consistent, and the numerical spectra follow the analytic expectation. I did not find a load-bearing error.\n\nWhat is actually new: the w ~ 1/gamma relation itself follows from known DBI kinematics, but the paper makes the observable consequences concrete. It numerically evolves the background and perturbations from inflation through radiation domination, computes the present-day GW spectra, and scans 10,000 {f0, kappa} pairs, keeping 835 that satisfy CMB, BBN, and LVK constraints and cross at least one future sensitivity curve. There is also a useful extra result in Appendix A: constant-roll DBI backgrounds with evolving sound speed can sustain a stiff phase for up to roughly 20 e-folds.\n\nWhat the paper does well: it keeps the canonical and DBI cases on the same footing, fixes the potential parameters to CMB-compatible values rather than fitting to the GW signal, and treats the exponential warp factor as a benchmark model, not as a best fit. The comparison with BBN, LVK, and the future experiment bands is standard and careful. The paper is honest about its own limitations: the frozen-mode transfer is approximate for UV modes and the affected part of the spectrum is dashed [b105-b107], and the full parameter scan is declared beyond scope [b177].\n\nThe soft spots are in proportion. The observable-signal claim rests on a single phenomenological input, rho_e^r from Eq. (71), which sets the stiff-phase duration and therefore which frequencies are enhanced. A concrete reheating mechanism, such as gravitational reheating, could tie rho_e^r to the DBI parameters themselves and shift or shrink the 835-point region. The authors explicitly leave the reheating mechanism unspecified, so this is a conditional-demonstration caveat rather than an internal error. The bigger practical problem is that no code or data is shipped, so the numerical spectra and the surviving parameter region cannot be independently audited. That is a fair thing to ask for in review. The neglect of g*(T) evolution is minor and the paper notes the effect is mostly an overall normalization.\n\nWho this is for: people working on post-inflationary stiff phases, GW phenomenology, and DBI inflation. It is not a paradigm shift, but it is a useful and honest piece of phenomenology. It deserves a serious referee. I would accept it for review and ask the authors to release the numerical code or at least a data table, and to add a sensitivity statement over rho_e^r. The central mechanism itself does not depend on that input, and I would cite the paper for the slope formula alone.","headline":"A clean, internally consistent demonstration that DBI kinetic domination softens the stiff-phase GW slope; the central physics holds up, and the main caveats are real but mostly acknowledged.","tokens_in":24689,"tokens_out":2679,"would_cite":true,"duration_ms":22775,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["98.80.Cq","04.30.-w"],"model":"deepseek/deepseek-v4-flash","headline":"After inflation, a Dirac–Born–Infeld scalar field lowers the stiff-phase equation of state below kination's w = 1, softening the gravitational-wave blue tilt and putting parts of the spectrum in reach of planned detectors.","keywords":["primordial gravitational waves","DBI inflation","kination","stiff equation of state","stochastic gravitational-wave background","non-canonical kinetic terms","reheating","post-inflationary cosmology"],"falsifier":"A measurement of the stochastic gravitational-wave background's spectral slope in the stiff-phase band would settle the claim: canonical kination predicts a slope of exactly 1, while this mechanism predicts a slope strictly between 0 and 1 for modes re-entering the horizon during the stiff phase, so a measured slope of 1 across the whole band would rule the softening out. A calculation-level test is to derive $\\rho_e^r$ from a specific reheating mechanism; if the resulting radiation abundance at the end of inflation exceeds the assumed $8.1 \\times 10^{-17} m_P^4$ by an order of magnitude or more, the stiff phase shortens and the predicted signals drop below the sensitivity curves shown in Fig. 5.","tokens_in":25190,"feed_emoji":"🌊","tokens_out":23942,"duration_ms":127522,"temperature":0.7,"pith_summary":"The paper aims to establish that the standard kination picture—a post-inflationary universe dominated by the kinetic energy of a scalar field, with equation-of-state parameter $w = 1$—is modified when the field's kinetic term has the Dirac–Born–Infeld (DBI) form: in that case $w(\\phi) \\simeq 1/\\gamma = \\sqrt{1 - 2f(\\phi)X}$, a value strictly below 1. The authors show that this softer equation of state changes the present-day gravitational-wave density spectrum in two competing ways: it weakens the blue tilt (the rise of $\\Omega_0^{\\rm GW}$ with frequency), but it also makes the field's energy density redshift more slowly, so the stiff phase lasts longer and imprints a broader band of frequencies. For an exponential warp factor $f(\\phi) = f_0 e^{\\kappa\\phi/m_P}$ they solve the background and perturbations numerically from inflation through reheating, finding 835 combinations of $\\{f_0, \\kappa\\}$ that satisfy CMB, big-bang-nucleosynthesis, and current LIGO–Virgo–KAGRA constraints while crossing the sensitivity curve of at least one planned experiment such as LISA, ET, DECIGO, or SKA. An appendix extends the analytic treatment to constant-roll DBI backgrounds with evolving sound speed, finding that stiff phases lasting up to roughly 20 e-folds with an expansion-averaged equation of state near 1/2 can be sustained there as well. If correct, the mechanism implies that gravitational-wave measurements could probe the kinetic structure of the inflaton's Lagrangian—information that observations of scalar perturbations alone cannot provide.","feed_headline":"Softening the stiff phase puts primordial gravitational waves in reach","feed_subtitle":"Lowering the stiff-phase equation of state below w = 1 changes the spectrum's slope and whether future detectors see it.","key_machinery":"The carrying mechanism is the DBI Lagrangian $P(X,\\phi) = (1/f(\\phi))(1 - \\sqrt{1 - 2f(\\phi)X}) - V(\\phi)$, whose Lorentz factor $\\gamma = 1/\\sqrt{1 - 2fX} = 1/c_s$ plays two roles at once: it sets the sound speed of the scalar fluctuations and, in the kinetically dominated limit, it becomes the equation of state, $w(\\phi) \\simeq 1/\\gamma$. The bridge from microphysics to observation is the spectral-slope formula $d\\log\\Omega_0^{\\rm GW}/d\\log k = n_t + 2(3w_k - 1)/(3w_k + 1)$, which turns any $w_k$ between 1/3 and 1 into a blue tilt between zero and one. To realize a sustained period of nearly constant $w$, the authors adopt the exponential warp factor $f(\\phi) = f_0 e^{\\kappa\\phi/m_P}$ with $\\kappa = \\sqrt{3(1/w + 1)}$, restricted to $\\sqrt{6} < \\kappa < \\sqrt{12}$ so that the stiff-phase condition $1/3 < w < 1$ holds; the amplitude $f_0$ then controls the duration of the phase through the separately conserved radiation density $\\rho_r(N) = \\rho_e^r e^{-4N}$, whose end-of-inflation value $\\rho_e^r$ fixes the timing of reheating. The verification is numerical: the Mukhanov–Sasaki and tensor mode equations are integrated on the full background, without the slow-roll approximation, from the start of inflation until deep into radiation domination.","core_discovery":"The central claim, stated by the authors in their setup and conclusions, is an identity for the post-inflationary stiff phase: once a DBI scalar field becomes kinetically dominated ($X \\gg V$), its equation-of-state parameter simplifies to $w(\\phi) \\simeq 1/\\gamma = \\sqrt{1 - 2f(\\phi)X}$, giving $1/3 < w < 1$ for $0 < fX < 4/9$ rather than the canonical kination value $w = 1$. Because the slope of the present-day gravitational-wave spectrum for modes re-entering the horizon during a stiff era is $d\\log\\Omega_0^{\\rm GW}/d\\log k = n_t + 2(3w_k - 1)/(3w_k + 1)$, this lowers the characteristic blue tilt from unity to a value between zero and one. At the same time, the slower redshift of the DBI field's energy density delays the onset of radiation domination, prolonging the stiff phase and widening the frequency band over which the softened tilt appears; the paper shows how this interplay can turn an unobservable canonical spectrum into a detectable one. The numerical extension verifies the analytic expectation with an exponential warp factor $f(\\phi) = f_0 e^{\\kappa\\phi/m_P}$ restricted to $\\sqrt{6} < \\kappa < \\sqrt{12}$, finding a surviving region spanning roughly ten orders of magnitude in $f_0$ and 835 parameter points whose spectra cross at least one future experiment's sensitivity curve. The high-frequency (UV) part of the computed spectra relies on the approximation that all tensor modes are frozen before horizon re-entry; the authors mark that region as less accurate and state that the uncertainty does not affect the detectability conclusions.","pith_inferences":["Inference: a measured slope $s$ in the stiff-phase band could be inverted to recover the stiff-phase equation of state through $w = (s+2)/(6-3s)$, and then, via $w \\simeq 1/\\gamma$, the warp factor $f(\\phi)$ at the relevant field values; this would connect the gravitational-wave band to the brane-tension scale of the warped geometry that motivates the DBI action.","Inference: the parameter combinations with observable signals have $c_s < 1$ near the end of inflation, so the model generically predicts enhanced small-scale non-Gaussianities and possibly primordial black holes or scalar-induced gravitational waves; a joint analysis of a GW detection with such small-scale probes would test the DBI origin independently of the background spectrum.","Inference: because the scan fixes the potential parameters and $\\rho_e^r$ to one benchmark, the 835-point region is a demonstration of viability rather than a full likelihood map; a scan over the inflationary parameters would probably shift the reachable region and could enlarge or shrink it.","Inference: the softening mechanism should extend to any non-canonical kinetic structure whose stiff-phase Lagrangian is not of the form $P \\propto X^n$; DBI is the concrete case treated here, so the paper effectively opens a family of softened stiff-phase spectra that future detectors could distinguish by their slopes."],"forward_implications":["A measured stochastic gravitational-wave background would carry a direct signature of the post-inflationary kinetic structure: a stiff-phase spectral slope strictly between 0 and 1 instead of the canonical kination slope of 1.","The combination of a weaker blue tilt and a longer stiff phase can move a spectrum that is undetectable in the canonical case into the sensitivity range of LISA, ET, DECIGO, or SKA; the paper exhibits 835 such parameter combinations at its benchmark potential.","The surviving parameter region spans roughly ten orders of magnitude in the warp-factor amplitude $f_0$, so the mechanism is not confined to a fine-tuned point in parameter space.","Because DBI and canonical dynamics are nearly indistinguishable at CMB scales but differ in the stiff-phase band of the gravitational-wave spectrum, GW observations could discriminate between the two where scalar measurements cannot.","The frequency band affected by the softened tilt is set by the assumed radiation abundance at the end of inflation, so the detectable window of Fig. 5 would shift or shrink under different reheating assumptions."],"supporting_citations":[{"why":"Supplies the DBI action whose non-canonical kinetic structure drives the softened equation of state.","marker":"[48, 49]"},{"why":"Supplies the runaway quintessential-inflation potential $V(\\phi) = V_0 e^{-\\lambda\\phi^n/m_P^n}$ used throughout the numerical evolution.","marker":"[59]"},{"why":"Defines the canonical kination baseline $w = 1$ that the DBI mechanism modifies during the stiff phase.","marker":"[13–15]"},{"why":"Establishes that modes re-entering during canonical kination produce a gravitational-wave spectrum with slope unity, the baseline this paper softens.","marker":"[16–18]"},{"why":"Source of the phenomenological value $\\rho_e^r = 8.1 \\times 10^{-17} m_P^4$ for the radiation density at the end of inflation, which sets the duration of the stiff phase.","marker":"[95]"},{"why":"Supplies the big-bang-nucleosynthesis bound on the integrated gravitational-wave energy density (as $\\Delta N_{\\rm eff}$) that the parameter scan must satisfy.","marker":"[96, 97]"},{"why":"Provides the power-law integrated sensitivity curves used to decide which spectra reach a future experiment's sensitivity band.","marker":"[80]"},{"why":"Supplies the CMB constraints on $n_s$, $A_s$, and $\\alpha_s$ that fix the benchmark potential and filter the parameter scan.","marker":"[5]"},{"why":"Supplies the upper bound on the tensor-to-scalar ratio used as an inflationary constraint in the scan.","marker":"[93]"}],"fun_headline_variants":["DBI kination softens the blue tilt of primordial gravitational waves","DBI lowers the stiff-phase equation of state to bring GWs within reach","A softer stiff phase broadens the gravitational-wave window for future detectors","DBI kination switches primordial gravitational waves from out-of-reach to observable","Lowering w below 1 softens the stiff-phase GW spectrum and its blue tilt"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing premise is that after inflation the universe contains only the kinetically dominated DBI scalar plus a separately conserved radiation fluid, with the radiation abundance at the end of inflation fixed to the phenomenological value $\\rho_e^r = 8.1 \\times 10^{-17} m_P^4$ (Eq. 71) and the reheating mechanism left unspecified—so if a realistic reheating process radiates earlier, sets a different $\\rho_e^r$, or adds extra energy components, the detectable region of parameter space shifts or shrinks.","fun_headline_variants_meta":{"raw":{"variants":["DBI kination softens the blue tilt of primordial gravitational waves","DBI lowers the stiff-phase equation of state to bring GWs within reach","A softer stiff phase broadens the gravitational-wave window for future detectors","DBI kination switches primordial gravitational waves from out-of-reach to observable","Lowering w below 1 softens the stiff-phase GW spectrum and its blue tilt"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.001034,"raw_usage":{"cost":0.0010338692,"is_byok":false,"cost_details":{"upstream_inference_cost":0.0010338692,"upstream_inference_prompt_cost":3.70552e-05,"upstream_inference_completions_cost":0.000996814},"total_tokens":6633,"prompt_tokens":1156,"completion_tokens":5477,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":936,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":5391}},"tokens_in":1156,"tokens_out":5477,"cost_source":"openrouter_usage","duration_ms":39139,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":5391,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-03T14:03:11.459261+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"falsifier":"A measurement of the stochastic gravitational-wave background's spectral slope in the stiff-phase band would settle the claim: canonical kination predicts a slope of exactly 1, while this mechanism predicts a slope strictly between 0 and 1 for modes re-entering the horizon during the stiff phase, so a measured slope of 1 across the whole band would rule the softening out. A calculation-level test is to derive $\\rho_e^r$ from a specific reheating mechanism; if the resulting radiation abundance at the end of inflation exceeds the assumed $8.1 \\times 10^{-17} m_P^4$ by an order of magnitude or more, the stiff phase shortens and the predicted signals drop below the sensitivity curves shown in Fig. 5.","supporting_citations":[{"cited_title":"The Atacama Cosmology Telescope: DR6 Power Spectra, Likelihoods and $\\Lambda$CDM Parameters","cited_arxiv_id":"2503.14452","evidence_quote":"Supplies the upper bound on the tensor-to-scalar ratio used as an inflationary constraint in the scan."},{"cited_title":"Inflation at the End of 2025: Constraints on $r$ and $n_s$ Using the Latest CMB and BAO Data","cited_arxiv_id":"2512.10613","evidence_quote":"Source of the phenomenological value $\\rho_e^r = 8.1 \\times 10^{-17} m_P^4$ for the radiation density at the end of inflation, which sets the duration of the stiff phase."},{"cited_title":"Planck constraints on single-field inflation","cited_arxiv_id":"1305.3044","evidence_quote":"Provides the power-law integrated sensitivity curves used to decide which spectra reach a future experiment's sensitivity band."}],"review_version":1}