{"id":"1b1b4117-0f6e-4e20-9e0a-d0b50f89b1ae","arxiv_id":"2608.11453","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Spin-fluid torsion cannot sustain inflation beyond about 0.46 e-folds, but a coupling-fed axial torsion condensate can hold an exact de Sitter background with an algebraic exit.","lead":"This paper computes how much cosmic inflation fermion spin can buy in Einstein-Cartan gravity: at most 0.46 e-folds, far short of the required 50 to 60. It then shows that a non-propagating axial torsion condensate, fed by a Gauss-Bonnet coupling, can hold an exact de Sitter background, but its perturbation spectrum is not yet computed.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Attractor eigenvalue checks out on an independent linearization; the load-bearing gap is the uncomputed perturbation sector, which the authors explicitly leave open.","rationale":"I read the central claim as the existence of an exact de Sitter background sustained by an algebraic torsion condensate in the axial branch, with a universal attractor eigenvalue and an algebraic exit, at background level. What must be true for that claim: the field equations admit (14)-(15); the constraints reduce the dynamics to a single first-order flow whose fixed point has eigenvalue -3H; the adiabatic promotion of U' is controlled; and the exit at U'_floor is consistent. I checked the solution algebra and found it consistent. I then independently linearized the scalar equation on the constraint surface at two benchmark solutions; in both cases F'(v) = -3H exactly, confirming Eq. (18) rather than refuting it. The adiabaticity ratio for the worked profile stays below 1e-2 and the e-fold count Ne = 0.81 phi_c is reproduced by direct integration of the slaved branch. Thus the reader's weakest-assumption concern about the underived attractor does not translate into a correctness risk in the tested instances. The remaining soft spot is the perturbation sector. The paper could not be clearer that this is open: the abstract says the branch was studied at background level and the perturbation spectrum remains open; Sec. 6 lists perturbations as the first of three open problems; and footnote 8 notes the axial fluctuation loses rank exactly at the snap. That last point is significant because the exit and handover are part of the proposed mechanism, so a perturbative breakdown there would compromise the physical story even if the background is smooth. This is a real, load-bearing limitation, but it is explicitly acknowledged and bounded, so the appropriate verdict is conditional rather than reject or unverified. My read does not change the reader's verdict; it sharpens the condition on which it rests. The concrete test is the deferred quadratic-action calculation, which would settle whether the axial branch is a viable inflationary mechanism or only a curiosity of the background sector.","tokens_in":12654,"tokens_out":21265,"duration_ms":187001,"concrete_test":"Compute the quadratic action for scalar and tensor perturbations on the axial-branch background after integrating out the algebraic torsion, following the route the authors propose. Then evaluate the kinetic matrix as a function of delta = U'/U'_floor - 1 down to the snap, and (i) check for ghosts and gradient instabilities in the sub-horizon limit; (ii) extract the scalar spectral index and tensor-to-scalar ratio for the worked profile (V0=17/2, U' = 1/16 [1 + (phi/phi_c)^2]^{-1}). If the kinetic determinant vanishes at delta = 0 or the spectrum is not nearly scale-invariant, the pseudo-inflation mechanism is not physically viable.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The background construction is internally consistent: the exact solution (14)-(15) satisfies the displayed field equations, and the constraint-surface reduction to a single flow p-dot = F(p) reproduces the claimed eigenvalue F'(v) = -3H. I verified this by direct linearization at two benchmarks (v=f0=1, V0=17/2 and v=3/2, f0=1, V0=93/8): in both cases the universal eigenvalue identity holds, so the reader's specific worry about Eq. (18) being wrong does not land in the tested cases. The adiabaticity ratio for the worked profile is also below 1e-2 along the branch, and the e-fold integral Ne = 0.81 phi_c reproduces. The real load-bearing gap is therefore the perturbation sector, which the paper itself declares open: 'Nothing in this work is a statement about the perturbations of the axial branch.' Footnote 8 adds that the algebraic determination of the axial fluctuation loses rank exactly at the snap, so the final stretch of the phase and the handover to the vectorial branch are uncontrolled at linear order. If the quadratic action reveals ghosts, gradient instabilities, or a spectrum incompatible with CMB observations, the pseudo-inflation phase cannot play the role of inflation even though the background is exact. This is a bounded, honestly stated limitation rather than an internal inconsistency, but it is precisely the condition on which the cosmological relevance of the axial branch rests.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies whether torsion can sustain a quasi-de Sitter phase in early-universe cosmology. It first derives a kinematic budget for the Einstein-Cartan spin-fluid bounce, showing that for any barotropic index w in [0,1] the accelerated window after the bounce is at most (1/3) ln 4 ≈ 0.46 e-folds, and that positive spatial curvature only shortens it. It then moves to Riemann-Cartan cosmology with a nonminimally coupled scalar, where torsion is sourced by couplings rather than by spin. The vectorial torsion branch is shown to reproduce Palatini inflation with its known observables. The axial branch, which requires a Gauss-Bonnet coupling, is the paper's main new result: the authors exhibit an exact de Sitter solution with constant axial torsion f0, constant scalar velocity v, linear coupling U = u1 phi, and flat potential, with the constants locked by Eqs. (14)-(15). They claim this solution is an attractor of the homogeneous dynamics with eigenvalue exactly -3H, that the phase exits when U' reaches the floor 5/(16 V0), and that effective-theory consistency places the episode near the Planck scale. The perturbation spectrum of the axial branch is explicitly left open.","tokens_in":12937,"tokens_out":6087,"duration_ms":64828,"significance":"If the background claims hold, the paper gives a clean negative result for the original spin-fluid torsion-inflation proposal, and a constructive two-branch alternative in which torsion is sustained by couplings. The budget bound in Eq. (3) is a genuinely parameter-free kinematic statement, and the exact solution in Eqs. (14)-(15) is a concrete, checkable contribution to Riemann-Cartan cosmology: an algebraic, non-propagating axial torsion condensate that supports an exact de Sitter background without slow-roll. The identification of the vectorial branch with Palatini inflation is useful and connects the framework to an observationally viable program. The paper is also commendably explicit about its limitations: it states that the perturbation spectrum of the axial branch is open and that the exit handover needs numerical treatment. The main significance gap is that the cosmological viability of 'pseudo-inflation' on the axial branch is precisely what remains unestablished: without a perturbative analysis, the background attractor cannot yet be connected to observable predictions.","major_comments":[{"comment":"The claim that the axial-branch homogeneous dynamics reduce to a single first-order flow p_dot = F(p) with F'(v) = -3H is load-bearing for the adiabatic tracking, the exit timing, and the e-fold estimate, but the function F(p) is never displayed and the eigenvalue is asserted rather than derived. As written, the reader cannot check that the reduction captures all directions on the constraint surface or that the eigenvalue is exactly -3H. Please derive F(p) explicitly and show the linearization at the fixed point, or move this derivation to an appendix.","section":"§5, Eq. (18)"},{"comment":"The manuscript explicitly leaves the perturbation spectrum of the axial branch open, stating that 'nothing in this work is a statement about the perturbations of the axial branch,' and the snap discussion notes that the algebraic determination of the axial fluctuation loses rank exactly at the exit. Since the paper presents this phase as 'pseudo-inflation,' the perturbation sector is not a peripheral issue: it is the condition under which the axial branch could play an inflationary role. I therefore ask either for a perturbative analysis (at minimum, the quadratic action and stability of the homogeneous background against inhomogeneous perturbations) or for a clear re-scoping of the abstract and title so that the claims are explicitly limited to the background solution, with the perturbation sector identified as a necessary precondition for the inflationary interpretation.","section":"§5 (after Eq. (17)) and §6"}],"minor_comments":[{"comment":"The sentence 'It is possible to verify that (14)-(15) solves all five equations' is vague; since the verification is claimed to be direct, please state that it follows by substitution or show the key algebra in one line.","section":"§5, after Eq. (15)"},{"comment":"The effective cutoff Lambda = 1/U' is introduced as 'naive,' yet it is used to pin the Hubble scale to H ≳ 0.3 in reduced Planck units. Please label this estimate explicitly as heuristic and clarify the dimensionful convention for U(phi) and Lambda, since the bound is presented as a 'sharp feature' of the mechanism.","section":"§5, Eq. (19)"},{"comment":"The main text says that positive spatial curvature shrinks the accelerated window, while footnote 1 notes that open curvature widens it. The qualification is important and should appear in the main text alongside Eq. (3), to avoid an apparent contradiction for readers who skip footnotes.","section":"§2, Eq. (3) and footnote 1"},{"comment":"The field equations (7)-(11) are imported from Ref. [15] and the companion paper [17]. A brief appendix or a more explicit mapping of the displayed equations to the numbering in Ref. [15] would make the paper more self-contained, given that several subsequent claims rest on this system.","section":"§1 and §3"},{"comment":"There are minor stylistic informalities, such as the parenthetical 'which nothing here forbids!' in the effective-theory discussion; these do not affect the substance but are out of place in a formal research paper.","section":"§5"}],"recommendation":"major_revision","confidential_remarks":"The background construction appears sound and honestly delimited, and the authors should be credited for stating the perturbation gap explicitly. My main reservation is scope: the paper's title and abstract make a cosmological claim ('pseudo-inflation'), but the axial branch's viability for that role depends on the perturbation sector, which is left open. The missing derivation of the attractor flow F(p) in §5 is also a concrete gap that should be filled before publication. If the authors re-scope the claims to background-level and add the derivation, or better, include a perturbative consistency analysis, the paper would be suitable. The heavy reliance on companion papers (Refs. [15,17,18]) is transparent and not, in itself, a problem."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version first. Two genuinely useful things here. One is a clean, parameter-free upper bound on the accelerated phase after the Einstein-Cartan spin-fluid bounce: at most (1/3) ln 4 ≈ 0.46 e-folds, for any barotropic index between dust and stiff matter, with positive curvature making it shorter. The other is an exact de Sitter background on the axial branch of the nonminimally coupled torsion cosmology, held up by a constant axial torsion condensate with no slow roll. Both survive contact with the algebra.\n\nThe budget bound is the cleanest part of the paper. Spin density dilutes as a^{-6}; the acceleration window between s^2=4ρ and s^2=(1+3w)ρ closes after the stated number. I re-derived it; it is correct. This is a genuinely useful negative result, and the paper states it without drama. The vectorial branch is honestly identified as Palatini inflation in torsional clothing, with its known observables, so that half of the paper is a translation, not a claim.\n\nThe axial solution (Eqs. 14-15) is new, as far as I can tell, and it does satisfy the displayed field equations. The one thing that worried me was Eq. (18), the attractor eigenvalue F'(v)=-3H, which the paper states as 'comes out' without showing the flow. I re-derived the single-flow reduction from the constraints and linearized at two benchmark points; the identity holds. The adiabatic tracking of a slowly varying U'(φ) is also controlled for the worked profile, and the e-fold integral reproduces.\n\nThe real load-bearing gap is perturbations. The authors are explicit: 'Nothing in this work is a statement about the perturbations of the axial branch.' That honesty matters, but it also means the phase is not yet an inflation mechanism. If the quadratic action has ghosts or gradient instabilities, or if the tensor spectrum is non-standard, the exact background does not rescue it. Footnote 8 adds that the axial fluctuation's algebraic determination loses rank exactly at the snap, so the final stretch and handover are uncontrolled at linear order. The paper deserves credit for saying this rather than burying it in prose.\n\nTwo lesser caveats. The operating point is forced near the Planck scale (H≳0.3 in reduced Planck units) to remain within the effective theory, and the authors flag the generic UV-operator problem themselves. The spin-fluid budget assumes the Weyssenhoff description at face value, but the conclusion is robust because no w rescues the window.\n\nCitation pattern is fine: the field equations come from Ref. [15], which is one author's earlier framework, but the axial solution and the budget bound are new. Self-citation here is contextual, not padding.\n\nWho is this for: torsion cosmologists and anyone cataloging alternatives to slow-roll inflation. It is a serious background-level contribution with one clean negative result and one new exact solution. It is not a complete inflationary model, and it does not pretend to be. Send it to peer review; a good referee will spend most of the time on perturbations, but the paper's positive results are real and its limitations are honestly declared.","headline":"The spin-fluid e-fold budget is settled with a clean 0.46 bound, and the axial-condensate de Sitter background is a new exact solution; the open perturbation sector is the real make-or-break.","tokens_in":13479,"tokens_out":6323,"would_cite":true,"duration_ms":84533,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A spin-fluid torsion bounce can accelerate expansion for at most 0.46 e-folds; an axial torsion condensate can sustain a quasi-de Sitter phase with an algebraic exit.","keywords":["torsion","Riemann-Cartan geometry","early-universe cosmology","Einstein-Cartan bounce","Gauss-Bonnet coupling","axial torsion condensate","pseudo-inflation","spin fluid"],"falsifier":"Compute the full linearized perturbation equations around the exact de Sitter solution (14)-(15) on the axial branch, including metric, scalar, and torsion fluctuations after eliminating the algebraic constraints. If any mode grows, or if the scalar power spectrum is far from scale-invariant, the pseudo-inflation phase is not viable; a simpler preliminary check is to derive the reduced flow $\\dot p = F(p)$ from Eqs. (7)-(11) and verify $F'(v)=-3H$ with no additional homogeneous directions.","tokens_in":12424,"feed_emoji":"🌀","tokens_out":16628,"duration_ms":137250,"temperature":0.7,"pith_summary":"This paper asks whether torsion can do the work of cosmic inflation. It first settles the budget of the classic spin-fluid mechanism: after the Einstein-Cartan bounce, a barotropic spin fluid with $w\\in[0,1]$ produces at most $\\frac{1}{3}\\ln 4\\simeq0.46$ e-folds of acceleration, since spin density dilutes as $a^{-6}$ and positive curvature only narrows the window. The paper then argues that in Riemann-Cartan cosmology with nonminimal couplings the source of torsion need not be spin. On the vectorial branch the theory reduces to the already viable connection-formulation of nonminimal inflation, while on the axial branch a linear Gauss-Bonnet coupling sustains an exact de Sitter background held by a constant axial torsion condensate. That condensate is a homogeneous attractor with eigenvalue $-3H$, ends when the coupling slope reaches the floor $5/(16V_0)$, and forces a near-Planckian Hubble rate, $H\\gtrsim0.3$; its perturbation spectrum is not yet computed.","feed_headline":"Torsion can't inflate alone; an axial condensate can","feed_subtitle":"Spin-fluid bounce yields at most 0.46 e-folds; a coupling-fed torsion condensate sustains de Sitter","key_machinery":"The load-bearing object is the axial torsion condensate on the axial branch of the nonminimally coupled Riemann-Cartan system. With metric and connection independent, the torsion tensor in a Friedmann-Lemaitre background reduces to one vector mode $h(t)$ and one axial mode $f(t)$; the axial mode survives only while the Gauss-Bonnet coupling satisfies $N=8U'\\dot\\phi Z$ (Eq. 12). On this branch, with $N=1$, $U=u_1\\phi$, and $V=V_0$, the constraints fix $Z$, $f_0$, $v$, $u_1$, and $H$ algebraically (Eqs. 14-15), so the torsion fluid acts as a cosmological constant, $\\rho_T=3H^2=-p_T$. The same constraint structure supplies the exit: $U'$ decreases with the roll speed and reaches the floor $5/(16V_0)$, where $f_0=0$ and the branch snaps back to the vectorial/no-axial branch.","core_discovery":"The paper's central claim is that torsion can sustain a quasi-de Sitter epoch, provided it is sourced by nonminimal couplings rather than by spin. In the Riemann-Cartan system of Eq. (4), the axial torsion mode $f$ survives only on the branch $N=8U'\\dot\\phi Z$; with $N=1$, $U=u_1\\phi$, and a flat potential $V=V_0$, the field equations admit the exact de Sitter background $a=e^{Ht}$, $\\dot\\phi=v$, $f=f_0$, with the constants locked by $Z^2=v^2+3f_0^2$, $H=(Z^2+f_0^2)/(2Z)$, $u_1=1/(8vZ)$, and $V_0=6f_0^2+\\tfrac52 v^2$ (Eq. 15). This condensate is a homogeneous attractor with eigenvalue exactly $-3H$ (Eq. 18), so nearby histories track it; as the coupling slope $U'$ declines with rising roll speed, $f_0^2$ drains to zero at the floor $U'_{\\rm floor}=5/(16V_0)$, and the branch snaps back to the torsionless scalar theory. The paper presents this as a background-level result: it is proven for homogeneous dynamics, the perturbation spectrum remains open, and effective-theory consistency requires $H\\gtrsim0.3$ in reduced Planck units.","pith_inferences":["Inference (beyond the paper): if the deferred perturbation calculation gives a nearly scale-invariant curvature spectrum, this becomes a distinct inflationary mechanism with no slow roll, and the parity-odd background $f_0\\neq0$ could imprint chiral correlations (TB and EB) in the microwave background.","Inference (beyond the paper): the two results suggest a natural two-stage historical sequence in one theory, with the spin-fluid bounce igniting expansion and the same nonminimal couplings then taking over as the condensate; the handover at the branch snap is a concrete numerical problem.","Inference (beyond the paper): because consistency forces a near-Planckian operating point, generic higher-curvature operators are likely unsuppressed; appending such operators to Eq. (4) and checking whether the exact de Sitter solution survives would be a direct robustness test.","Inference (beyond the paper): the branch-snap exit is a constraint-driven transition rather than a potential-driven one, so its gravitational-wave and reheating signatures could distinguish the condensate from ordinary slow-roll inflation once perturbations are computed."],"forward_implications":["The spin-fluid Einstein-Cartan bounce cannot by itself provide the needed 50 to 60 e-folds; the accelerated window is at most $\\frac13\\ln4\\simeq0.46$ e-folds for any barotropic index $w\\in[0,1]$, so the original torsion-inflation mechanism must be augmented.","The vectorial branch of coupling-sourced torsion is equivalent to the known connection-formulation of nonminimal scalar inflation, so its observational predictions already inherit a viable status.","The axial condensate gives a proof of principle that torsion can sustain a quasi-de Sitter background without slow roll, with an exactly flat potential and an exit written into the constraint structure rather than into the potential.","Any successful use of the axial branch operates near the Planck scale: the cutoff $\\Lambda=1/U'$ together with the branch condition forces $H\\gtrsim0.3$ in reduced Planck units, a narrow corridor.","The paper's claims are background-level; the perturbation spectrum, stability, and reheating history of the axial branch remain open, so the mechanism is not yet established as a complete cosmology."],"supporting_citations":[{"why":"launches the spin-dominated torsion inflation idea whose budget this paper computes","marker":"[1]"},{"why":"proposes torsion as an alternative to cosmic inflation and is the target of the e-fold budget","marker":"[2]"},{"why":"establishes the nonsingular bounce cosmology that yields the accelerated window","marker":"[3]"},{"why":"records that macroscopic spin fluids do not give a scale-invariant spectrum, the perturbation caveat behind the axial branch's open status","marker":"[6]"},{"why":"supplies the nonminimally coupled scalar-torsion action and complete field equations from which both branches are derived","marker":"[15]"},{"why":"introduces scalar-Euler-form coupling in Riemann-Cartan geometry, the seed of coupling-sourced torsion","marker":"[16]"},{"why":"shows Einstein-Cartan Higgs inflation lands on the connection-formulation side whose observables the vectorial branch imports","marker":"[19]"},{"why":"parallel derivation that integrating out algebraic torsion yields the metric-to-connection family, used to identify the vectorial branch","marker":"[23]"},{"why":"fixes the metric-versus-connection difference that the vectorial branch reproduces","marker":"[24]"}],"fun_headline_variants":["Spin torsion yields 0.46 e-folds; coupling-driven condensate yields de Sitter","Axial torsion condensate gives exact de Sitter; spin torsion falls short","Torsion needs nonminimal couplings, not spin, for sustained inflation","Torsion's condensate beats its spin: from 0.46 e-folds to de Sitter"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the claim, stated without derivation around Eq. (18), that on the axial branch all homogeneous histories are pulled toward the condensate at a single rate set by $3H$; if that reduction is wrong or another direction grows, the tracking picture, the exit timing, and the e-fold count all collapse.","fun_headline_variants_meta":{"raw":{"variants":["Spin torsion yields 0.46 e-folds; coupling-driven condensate yields de Sitter","Axial torsion condensate gives exact de Sitter; spin torsion falls short","Torsion needs nonminimal couplings, not spin, for sustained inflation","Torsion's condensate beats its spin: from 0.46 e-folds to de Sitter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000581,"raw_usage":{"total_tokens":2889,"prompt_tokens":1251,"completion_tokens":1638,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":867,"completion_tokens_details":{"reasoning_tokens":1544}},"tokens_in":867,"tokens_out":1638,"duration_ms":13463,"temperature":1.0,"reasoning_tokens":1544,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:12:45.536411+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full linearized perturbation equations around the exact de Sitter solution (14)-(15) on the axial branch, including metric, scalar, and torsion fluctuations after eliminating the algebraic constraints. If any mode grows, or if the scalar power spectrum is far from scale-invariant, the pseudo-inflation phase is not viable; a simpler preliminary check is to derive the reduced flow $\\dot p = F(p)$ from Eqs. (7)-(11) and verify $F'(v)=-3H$ with no additional homogeneous directions.","supporting_citations":[{"cited_title":"Spin-dominated inflation in the Einstein-Cartan theory,","cited_arxiv_id":null,"evidence_quote":"launches the spin-dominated torsion inflation idea whose budget this paper computes"},{"cited_title":"Thermal fluctuations in Einstein-Cartan-Sciama-Kibble-Dirac bouncing cosmology","cited_arxiv_id":"1201.0316","evidence_quote":"records that macroscopic spin fluids do not give a scale-invariant spectrum, the perturbation caveat behind the axial branch's open status"}],"review_version":1}