{"id":"88165fca-85ec-4160-ad79-0cee9a40b2ea","arxiv_id":"2603.17654","paper_version":2,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Non-polynomial quasi-topological gravities in d=4 yield three singularity-free cosmologies (de Sitter origin, bounce, or Minkowski loitering) for radiation and standard matter while recovering the infrared Friedmann limit.","lead":"Modified gravity theories that already fix black-hole singularities can also erase the Big Bang in four dimensions. Three generic early-universe histories emerge—de Sitter start, bounce, or eternal loitering—without needing exotic matter.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"Multi-valued Lagrangian (bounce) and infinite-affine density divergence (de Sitter origin) leave the non-singularity claim under-specified; manuscript encoding also blocks verification of the modified Friedmann equations.","rationale":"The reader correctly identified the load-bearing soft spots: physical acceptability of multi-valued Lagrangians and of density divergence only at infinite affine distance, plus the inability to verify IR recovery and second-order structure from the corrupted text. Those are precisely the conditions that must hold for the three-scenario claim to stand. The third (asymptotically Minkowski / loitering) family looks more robust on the abstract’s wording, but the paper markets three generic families, so the weaker two control the claim. No stronger verdict shift is warranted: the program (quasi-topological regularization of singularities) is established, the abstract is carefully hedged, and the issues are fixable with explicit actions, branch analysis, and a readable manuscript rather than internal contradiction. Hence CONDITIONAL remains appropriate; confidence stays low until the encoding is fixed and the two checks above are passed. No ethics or data issues apply.","tokens_in":3844,"tokens_out":690,"duration_ms":25132,"concrete_test":"Recover or re-derive the explicit algebraic relation between the effective Friedmann function and the non-polynomial curvature combination used for the bounce; restrict to one continuous branch through the would-be bounce and check that the modified Friedmann equation remains single-valued and second-order with continuous H(t), Ḣ(t). Separately, integrate past-directed geodesics for the de Sitter-origin solution and confirm that every curvature invariant stays bounded while ρ→∞ only as affine parameter →∞; if either check fails, reclassify that scenario.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract’s claim of three generic non-singular cosmologies with correct IR limit rests on two constructions whose physical status is not secured by the abstract alone, and the supplied full text is encoding-corrupted (mojibake) so the derivations cannot be checked. (1) Bounce: the paper states it “requires a multi-valued Lagrangian.” A multi-valued action density is not a standard classical field theory—the variational principle and the map from curvature to the effective Friedmann function may fail to be single-valued, branch selection is unspecified, and well-posedness of the EOM through the bounce is not guaranteed. Without a continuous single-valued physical branch with continuous H and Ḣ, this is not a demonstrated non-singular cosmology in a well-defined theory. (2) de Sitter origin: curvature invariants remain finite while matter density diverges at infinite affine distance. Geodesic completeness may hold, but infinite energy density is still a matter-sector pathology; whether this counts as “non-singular” depends on a narrow geometric criterion that must be justified against standard singularity notions. The Minkowski/loitering case is cleaner (non-singular, sub-Planckian density for the past eternity), but the headline claim of three generic families is only as strong as the weaker two. IR recovery of Einstein gravity and second-order character of the full non-polynomial series are asserted as setup but cannot be verified from the unreadable manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript studies whether non-polynomial quasi-topological gravities in four dimensions, already known to resolve black-hole singularities, can also remove cosmological Big-Bang singularities. From the abstract and stated setup, the authors derive modified Friedmann equations for these theories and classify three generic families of solutions that recover the standard infrared (late-time) Friedmann limit for pure radiation or other standard matter: (i) emergence from a de Sitter phase with finite curvature invariants but a matter density that diverges only at infinite affine distance; (ii) a bouncing cosmology that requires a multi-valued Lagrangian; and (iii) an asymptotically Minkowski (eternally loitering) origin whose matter density remains non-singular, approximately constant, and sub-Planckian for the entire past. The central claim is that the same quasi-topological mechanism that regularizes black holes yields non-singular cosmologies with correct IR behavior.","tokens_in":4171,"tokens_out":1139,"duration_ms":16921,"significance":"If the derivations hold, the work would extend the singularity-resolution programme of quasi-topological gravity from static black holes to dynamical FLRW cosmologies in d=4, a setting of direct observational interest. The three-scenario classification, the emphasis on infrared recovery of Einstein gravity, and the identification of a clean loitering branch with sub-Planckian density are potentially useful contributions to modified-gravity cosmology. The result would be of interest to the community working on higher-curvature and quasi-topological theories, provided the constructions are shown to be well-defined classical field theories and the modified Friedmann equations are fully specified and checked.","major_comments":[{"comment":"The bouncing scenario is stated to require a multi-valued Lagrangian. A multi-valued action density is not a standard classical field theory: the variational principle, the map from curvature to the effective Friedmann function, and branch selection through the bounce are not automatically well-defined. The manuscript must either (a) exhibit a continuous single-valued physical branch with continuous H and Ḣ across the bounce and a well-posed initial-value problem, or (b) justify why a multi-valued Lagrangian remains an acceptable classical description. Without that, the bounce cannot be counted as a demonstrated non-singular cosmology in a well-defined theory.","section":null},{"comment":"For the de Sitter-origin scenario the paper counts the solution as non-singular because curvature invariants remain finite while the matter density diverges only at infinite affine distance. Infinite energy density is still a matter-sector pathology. The manuscript needs an explicit comparison with standard singularity criteria (geodesic incompleteness, curvature blow-up, and matter-sector regularity) and a clear statement of which criterion is being adopted and why a density divergence at infinite affine distance is acceptable. Otherwise the non-singularity claim for this branch rests on a narrow geometric criterion that is not yet justified.","section":null},{"comment":"The infrared recovery of Einstein gravity and the second-order character of the modified Friedmann equations are asserted for the full non-polynomial curvature series. These properties are load-bearing for all three scenarios. The manuscript should display the explicit form of the modified Friedmann equations (or the effective algebraic relation between H^{2} and ρ), state the conditions under which the series reduces to GR at low curvature, and confirm that no higher-than-second-order derivatives of the scale factor appear for the FLRW ansatz used. Without that verification the classification of solutions cannot be trusted.","section":null},{"comment":"The third (asymptotically Minkowski / loitering) branch is the cleanest of the three: density non-singular and sub-Planckian for the past eternity. The paper should quantify how generic this branch is within the space of non-polynomial quasi-topological couplings, and whether it requires fine-tuning relative to the de Sitter-origin and bounce branches. A parameter-space sketch or existence theorem would strengthen the claim that three generic families exist.","section":null}],"minor_comments":[{"comment":"The abstract is clear, but the body should define notation for the non-polynomial quasi-topological densities and the precise sense of “correct infrared limit” (e.g., leading Einstein–Hilbert term plus controlled higher-curvature corrections) before the cosmological analysis.","section":null},{"comment":"When discussing affine distance and geodesic completeness for the de Sitter-origin case, a short appendix computing the affine parameter for radial null and timelike geodesics would make the “infinite affine distance” claim self-contained.","section":null},{"comment":"References to the black-hole singularity-resolution results in quasi-topological gravity should be cited with equation or theorem numbers so that the analogy to the cosmological case is easy to check.","section":null}],"recommendation":"major_revision","confidential_remarks":"The review package supplied a full-text block that is encoding-corrupted (mojibake) and unreadable; only the abstract is usable. I have therefore assessed the load-bearing claims as they appear in the abstract and the stated setup, and flagged the multi-valued Lagrangian and density-divergence issues as requiring major revision. Once a clean PDF is available the modified Friedmann equations and regularity proofs should be re-checked line by line. Scope fit for a gr-qc journal is appropriate if the technical gaps are closed."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The punchline is simple: they take the quasi-topological machinery that already resolves black-hole singularities and ask whether the same higher-curvature setup can kill the Big Bang for ordinary matter in four dimensions. From the abstract they get three families with the right late-time Friedmann limit—emergence from de Sitter, a bounce, and an asymptotically Minkowski loitering phase.\n\nWhat is actually new is the cosmological classification itself, not the gravity theory. Quasi-topological gravities and their BH regularity are established; cataloguing the three IR-safe, singularity-free cosmologies for radiation (and standard matter) is a legitimate extension. The Minkowski/loitering branch looks cleanest: density stays finite, sub-Planckian, and roughly constant into the past. That is worth having on the table.\n\nSoft spots are real and uneven. The bounce explicitly needs a multi-valued Lagrangian. That is not a minor technicality—variational principle, branch choice, and well-posedness of the equations through the bounce are all open until they show a continuous physical branch with continuous H and Ḣ. The de Sitter-origin case keeps curvature invariants finite but lets matter density diverge at infinite affine distance; geodesic completeness may hold, yet infinite energy density is still a matter-sector pathology, and calling it “non-singular” needs a clear criterion. The full manuscript we were given is encoding-corrupted, so the modified Friedmann equations, the explicit non-polynomial series, and the regularity proofs cannot be checked. That is a verification problem, not a proof of error.\n\nThis is for people already working on higher-curvature gravity, singularity theorems, and early-universe alternatives to inflation. A serious referee should see it: the claim shape is falsifiable and sits inside an active program. I would not cite it yet until the multi-valued and infinite-density issues are cleaned up and a readable text is available. Send it to peer review; do not desk-reject.","headline":"Abstract-only readable claim of three non-singular cosmologies in d=4 quasi-topological gravity; bounce and de Sitter-origin cases rest on constructions that need hard scrutiny.","tokens_in":4763,"tokens_out":506,"would_cite":false,"duration_ms":6644,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.50.Kd","98.80.Jk","04.20.Dw"],"model":"grok-4.5","headline":"Four-dimensional non-polynomial quasi-topological gravity yields three families of cosmologies without a Big Bang, all recovering the usual late-time expansion.","keywords":["quasi-topological gravity","non-polynomial curvature","modified Friedmann equations","non-singular cosmology","Big Bang singularity","bouncing universe","de Sitter origin","loitering universe"],"falsifier":"Derive the full modified Friedmann equation for a concrete non-polynomial quasi-topological series and check whether any of the three claimed early-time branches (de Sitter origin, bounce, Minkowski loitering) actually solves it for radiation while recovering the standard late-time expansion.","tokens_in":4696,"feed_emoji":"🌌","tokens_out":637,"duration_ms":4906,"temperature":0.7,"pith_summary":"The same higher-curvature quasi-topological theories already known to smooth black-hole interiors can also eliminate the Big Bang. In four dimensions, with non-polynomial curvature terms, the modified Friedmann equations admit three generic early-universe solutions for ordinary matter (including pure radiation). One universe emerges from a de Sitter phase whose curvature stays finite, even though the matter density formally diverges only at infinite affine distance. A second solution bounces, but needs a multi-valued Lagrangian. The third begins asymptotically from Minkowski space, with a nearly constant, sub-Planckian density that never becomes singular. All three recover ordinary Einstein gravity at late times, so the singularity resolution is an ultraviolet modification that leaves the observed cosmology intact.","feed_headline":"Three ways gravity can kill the Big Bang","feed_subtitle":"Quasi-topological corrections yield de Sitter birth, bounce, or eternal loitering without spoiling late cosmology","key_machinery":"The modified Friedmann equations obtained by reducing non-polynomial quasi-topological curvature densities to a second-order equation for the scale factor; these equations retain the correct Einstein infrared limit while opening new ultraviolet branches that avoid a=0.","core_discovery":"In d=4 non-polynomial quasi-topological gravities the modified Friedmann equations possess three families of solutions that match the standard infrared cosmology yet never encounter a Big-Bang singularity for radiation or other ordinary matter: (i) emergence from de Sitter with finite curvature invariants, (ii) a bounce that requires a multi-valued Lagrangian, and (iii) an eternally loitering, asymptotically Minkowski origin whose density remains finite and sub-Planckian for the entire past.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Three non-singular cosmologies from quasi-topological gravity","Gravity yields de Sitter birth bounce or eternal loitering","Modified Friedmann equations erase Big Bang singularities","Quasi-topological terms allow singularity-free radiation universes","Non-polynomial gravity produces three Big Bang free origins"],"cache_read_input_tokens":128,"weakest_assumption_plain":"That the non-polynomial quasi-topological Lagrangians remain physically acceptable when they are multi-valued or when they allow a density that diverges only at infinite past affine distance.","fun_headline_variants_meta":{"raw":{"variants":["Three non-singular cosmologies from quasi-topological gravity","Gravity yields de Sitter birth bounce or eternal loitering","Modified Friedmann equations erase Big Bang singularities","Quasi-topological terms allow singularity-free radiation universes","Non-polynomial gravity produces three Big Bang free origins"]},"model":"grok-4.5","effort":"low","cost_usd":0.006844,"raw_usage":{"total_tokens":1668,"prompt_tokens":743,"num_sources_used":0,"completion_tokens":62,"cost_in_usd_ticks":68440000,"prompt_tokens_details":{"text_tokens":743,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":863,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":743,"tokens_out":62,"duration_ms":6758,"temperature":1.0,"reasoning_tokens":863,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-15T11:55:53.430600+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Derive the full modified Friedmann equation for a concrete non-polynomial quasi-topological series and check whether any of the three claimed early-time branches (de Sitter origin, bounce, Minkowski loitering) actually solves it for radiation while recovering the standard late-time expansion.","supporting_citations":[],"review_version":1}