{"id":"d244a438-9c27-4527-b5b3-c2d7a170d5fd","arxiv_id":"2412.17896","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Hadronic axion models with d=6 and d=7 heavy-quark decay operators can produce two new cosmologically viable 'islands' at f_a ~ 10^12 and 10^14 GeV, extending the post-inflationary axion search window.","lead":"This paper extends the catalogue of hadronic QCD axion models to include heavy exotic quarks whose late decays can trigger a brief early matter-dominated era, opening new allowed mass ranges for the axion. The resulting 'islands' at axion decay constants near 10^12 and 10^14 GeV give axion searches new target regions beyond the standard post-inflationary window.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The d=7 island is not excluded by the reheating bound the reader invokes (T_RH,max is (H_I m_P)^{1/2}, not T_GH); the load-bearing uncertainty is instead the neglected string/domain-wall DM contribution, which the paper admits may erase both islands.","rationale":"The paper is honest and well-documented, and the reader's CONDITIONAL verdict is reasonable. However, the reader's stated weakest assumption is not the correct one: the post-inflationary requirement is T_RH > f_a, and T_RH can be as high as (H_I m_P)^{1/2} ~ 10^16 GeV, not T_GH ~ 6×10^12 GeV. So the d=7 island is not excluded by reheating; the paper's own caveat in Section 4.4 appears overly conservative or even erroneous. The genuine load-bearing concern is the neglected topological-defect DM contribution, which is a known, unquantified uncertainty. The paper flags it clearly and provides its catalogue for recalculation, which is good practice. The islands are realignment-only estimates; their 'viability' is conditional on defects being subdominant. This justifies the CONDITIONAL verdict, but for a different reason than the reader's T_GH objection. Hence no change to the verdict.","tokens_in":33391,"tokens_out":24747,"duration_ms":236163,"concrete_test":"For each f_a in the d=6 (0.25–2.0×10^12 GeV) and d=7 (1.0–1.7×10^14 GeV) islands, add the topological-defect abundance Ω_strings+DW from recent lattice simulations (e.g., refs [37], [128], [129]), scaled to the paper's EMD cosmology using the public code, or alternatively adopt the conservative f_a < 2×10^10 GeV limit from ref [129]. Recompute Ω_total h^2 = Ω_realignment + Ω_defects and test whether any catalogue model remains below 0.12. If none survive, the islands are not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of two viable post-inflationary islands is conditional on the assumption that the axion relic density is dominated by realignment and that the string/domain-wall contribution can be ignored. The islands' upper f_a boundaries are fixed by Ω_a h^2 < 0.12; any extra DM abundance from topological defects tightens this bound. Recent simulations (refs [37,128,129]) suggest the standard-cosmology upper bound may drop from ~2×10^11 GeV to ~10^10–10^11 GeV, and the paper itself states 'both the d=7 and the d=6 regions may disappear altogether' (Section 4.4). Since no quantitative estimate or even an upper bound is provided for Ω_strings+DW in the EMD scenario, the existence of the islands is not established; it is a working assumption. I also note that the reader's specific rejection of the d=7 island via T_GH is not correct: the maximum reheating temperature after inflation is T_RH,max ≈ (H_I m_P)^{1/2} ≈ 10^16 GeV for H_I ≈ 4×10^13 GeV, well above f_a ≈ 10^14 GeV. The Gibbons–Hawking temperature is the de Sitter vacuum temperature, not the reheating temperature. Thus the d=7 island is not internally ruled out; the unresolved topological-defect contribution is the real load-bearing caveat.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper extends the catalogue of KSVZ hadronic QCD axion models to include heavy fermion representations compatible with higher-dimensional decay operators up to d≤9. Using DECO for operator enumeration and the Landau-pole criterion as a preselection, the authors identify new Q representations and solve coupled Boltzmann equations that allow multiple Qs to induce a period of early matter domination. They compute the axion realignment abundance with MiMeS and derive constraints from BBN (via the proxy w(T_BBN)>0.3), dark matter overproduction, and hot DM. The main results are an updated axion-photon coupling band whose central region is approximately f_a-independent, and two post-inflationary 'islands' around f_a∼10^12 GeV (d=6) and f_a∼10^14 GeV (d=7). The paper includes extensive supplementary material and is unusually explicit about its assumptions and limitations.","tokens_in":33730,"tokens_out":8652,"duration_ms":80424,"significance":"If the islands survive scrutiny, they provide concrete haloscope targets at ma∼10 μeV and beyond the standard post-inflationary window, tying a possible axion discovery to early matter domination and to the scale Λ_EFT. The paper's strengths are the exhaustive operator enumeration with the public DECO-based pipeline, the public catalogues and analysis scripts, and the stable central band that is robust to many model-building choices. The headline islands, however, are conditional on neglecting the topological-defect contribution to Ω_a h^2 and on the choice Λ_EFT=M_P; the manuscript itself states that the islands could disappear under stronger defect contributions. The significance is therefore moderate and depends on how these caveats are resolved.","major_comments":[{"comment":"The upper boundaries of both islands are set by the condition Ω_a h^2 < 0.12, but only realignment production is included in the computation. The paper itself notes that 'both the d=7 and the d=6 regions may disappear altogether' once the cosmic-string/domain-wall contribution is included (Section 4.4), and the cited simulations suggest the standard-cosmology upper bound on f_a could drop by O(2)–O(200). Because no quantitative estimate, not even a conservative upper bound, is provided for Ω_strings+DW in the EMD scenario, the existence of the islands is not established by the present analysis. I recommend either adding a quantitative defect estimate, even one with large uncertainties, or explicitly reclassifying the islands as benchmark regions conditional on defect-neglect in the abstract and conclusions.","section":"4.4 (with Eq. (3.2) and Eq. (3.10))"},{"comment":"The island structure is strongly sensitive to the choice Λ_EFT=M_P in Eq. (3.3). In Section 4.4 the authors state that lowering Λ_EFT to about 0.7M_P merges the d=6 island with the standard region, Λ_EFT∼0.01M_P merges the d=7 island into the d=6 region, and d=8 operators become viable for Λ_EFT≲0.1M_P. Since no theoretical argument uniquely selects Λ_EFT=M_P, the two-island claim is a benchmark-dependent result rather than a robust prediction. The main text should either scan Λ_EFT or present the islands with an explicit confidence statement tied to the cutoff choice.","section":"3.2.1 and 4.4"},{"comment":"The manuscript compares the Gibbons–Hawking temperature T_GH=H_I/2π with f_a and uses this to suggest that the 'limit on the maximum possible reheating temperature... may rule out the post-inflationary scenario for the d=7 island' (Section 4.4). This is not the correct criterion: the maximum reheating temperature after inflation is of order T_RH,max∼(H_I m_P)^{1/2}≈10^16 GeV for H_I≈4×10^13 GeV, which is well above f_a∼10^14 GeV. The de Sitter vacuum temperature T_GH is not the reheating temperature and does not by itself preclude PQ symmetry restoration after inflation. The d=7 island is therefore not internally ruled out by this argument; please correct the text and remove or replace the T_GH-based caveat.","section":"3.2.1 and 4.4"}],"minor_comments":[{"comment":"The sentence 'shows a period of EDM' should read 'EMD'.","section":"3.2.1"},{"comment":"The phrase 'neglected in in ref. [19, Fig. 4]' contains a duplicated 'in'.","section":"4.2"},{"comment":"The string 'TBBN = 1 MeVand' is missing a space and should read 'MeV and'.","section":"3.2.1"},{"comment":"The neglect of entropy injection from Q annihilations and inverse decays is stated but not quantitatively justified; a short estimate of the size of this effect would make the Boltzmann treatment more robust.","section":"3.2.1 (Eq. (3.2))"},{"comment":"The BBN consistency proxy w(T_BBN)>0.3 is a useful first cut, but since the lower f_a boundaries of the islands are set by this criterion, a dedicated BBN calculation (e.g., with AlterBBN or ACROPOLIS) would strengthen the quantitative boundaries.","section":"3.2.1 (Eq. (3.8))"}],"recommendation":"major_revision","confidential_remarks":"This is a solid phenomenological catalogue paper with reproducible code and transparent caveats. My main concern is that the abstract's 'two new viable model islands' overstates robustness given the acknowledged topological-defect and Λ_EFT sensitivities; I would advise the authors to condition the headline claim explicitly on defect-neglect and on the benchmark Λ_EFT=M_P. The T_GH point should be corrected before publication, but it does not invalidate the core calculation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a careful, reproducible catalogue paper, and the two new islands are worth taking seriously as possibilities—but the load-bearing uncertainty is the neglected string/domain-wall contribution, not the reheating bound the reader uses to dismiss the d=7 island.\n\nWhat's new: the complete Landau-pole-allowed representation list up to d=9, the multi-heavy-quark early-matter-domination treatment via dimensional signatures, and the f_a-dependent axion band. The central band is a genuinely useful update and looks robust. The paper ships code and data, and it is unusually candid about its own limitations—Section 4.4 names the threshold choices, the BBN proxy, confinement, and the topological-defect problem, including the statement that both islands may disappear. That honesty is real credit.\n\nWhere I part with the reader: the T_GH argument against the d=7 island is not right. The relevant maximum reheating temperature is T_RH,max ~ (H_I m_P)^{1/2} ~ 10^16 GeV, far above f_a ~ 10^14 GeV; the Gibbons-Hawking temperature is not the reheating temperature. The stress-test note is correct here. The paper's own sentence about the bound 'may rule out' is more caution than an internal inconsistency.\n\nThe soft spots are real but not fatal to the paper's core value. The islands' boundaries depend on Lambda_EFT = M_P, on the w(T_BBN)>0.3 proxy, on m_Q=f_a, and on ignoring topological defects. The last is the big one: recent string simulations suggest standard-cosmology f_a upper bounds drop by O(2)-O(200), and the paper admits the islands may disappear. So the headline claim should be read as \"viable if the realignment-only assumption holds,\" not as a firm prediction. The central band is less exposed to these caveats, which is what makes the paper worth using.\n\nI don't see a circularity problem: theta_eff=2.2 comes from standard-cosmology dark matter conditions, and the islands are outputs of the Boltzmann equations, not fitted inputs.\n\nWho it's for: axion phenomenologists setting search targets, and anyone working on nonstandard cosmological histories. It deserves a serious referee; the work is formal enough and reproducible enough that referee time is well spent, even if the islands get revised down.","headline":"A careful, reproducible extension of the KSVZ axion catalogue whose two new model islands are real possibilities, not established targets: the load-bearing caveat is the neglected topological-defect contribution, not the reheating bound the reader uses.","tokens_in":34302,"tokens_out":2290,"would_cite":true,"duration_ms":24313,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper identifies two new cosmologically viable mass windows for hadronic QCD axion dark matter, at f_a ~ 10^12 GeV and f_a ~ 10^14 GeV.","keywords":["QCD axion","hadronic axion model","KSVZ model","early matter domination","axion dark matter","axion-photon coupling","model catalogue","Peccei-Quinn symmetry"],"falsifier":"Measure the tensor-to-scalar ratio $r$ at the level of about 0.001, which would fix the inflationary Hubble scale at $H_I \\approx 6\\times 10^{12}$ GeV and the Gibbons-Hawking temperature at $T_{\\mathrm{GH}} \\approx 10^{12}$ GeV; with $T_{\\mathrm{GH}}$ inside the $d=6$ island's range ($f_a$ up to $2\\times 10^{12}$ GeV), the post-inflationary interpretation for that island would be ruled out, eliminating the paper's main new prediction.","tokens_in":33155,"feed_emoji":"🎯","tokens_out":10408,"duration_ms":78653,"temperature":0.7,"pith_summary":"The paper extends the catalogue of hadronic (KSVZ) QCD axion models to include heavy, coloured fermions that decay through higher-dimensional operators, up to dimension 9. Such operators make the fermions decay slowly enough to drive a phase of early matter domination, which relaxes the usual lifetime bounds and opens up cosmological viability. The authors find two new 'model islands' of viable axion decay constant $f_a$ in the post-inflationary scenario: a $d=6$ island at $f_a \\sim 10^{12}$ GeV and a $d=7$ island at $f_a \\sim 10^{14}$ GeV, beyond the standard post-inflationary mass region. These islands give concrete targets for axion searches.","feed_headline":"Two new mass windows open for hadronic axion dark matter","feed_subtitle":"The d=6 and d=7 decay operators put viable axion dark matter at fa ~ 10^12 and 10^14 GeV, beyond the standard window.","key_machinery":"The central object is the lowest-dimension decay operator for each heavy Peccei-Quinn fermion $Q$, since its dimension $d$ sets the decay rate $\\Gamma \\sim m_Q (m_Q/\\Lambda_{\\mathrm{EFT}})^{2(d-4)}$ and hence controls whether the fermions freeze out, drive an early matter domination phase, and decay before Big Bang nucleosynthesis. Models are labelled by their 'dimensional signature'—the multiplicities of the lowest decay-operator dimension for each $Q$—because the cosmology of multiple $Q$s is essentially determined by the largest $d$ among the lowest-dimensional operators. The analysis combines Boltzmann equations for the temperature and fermion number densities with the axion realignment equation solved by the code MiMeS, using the temperature-dependent axion mass from lattice QCD.","core_discovery":"The paper claims that, in the post-inflationary Peccei-Quinn symmetry breaking scenario, hadronic axion models with heavy fermions decaying via dimension-6 and dimension-7 operators form two cosmologically self-consistent 'model islands': the $d=6$ island spanning $f_a \\in [0.25, 2.0]\\times 10^{12}$ GeV and the $d=7$ island spanning $f_a \\in [1.0, 1.7]\\times 10^{14}$ GeV. The islands are bounded from below by requiring that the fermions decay before Big Bang nucleosynthesis and from above by the dark matter relic density constraint. The paper also updates the hadronic axion band for the axion-photon coupling and finds that its central region is nearly independent of $f_a$, whereas the maximal coupling grows with $f_a$ because the Landau pole criterion becomes less restrictive at larger $f_a$.","pith_inferences":["If the Gibbons-Hawking bound on the post-inflationary scenario is enforced strictly, the $d=7$ island at $f_a \\sim 10^{14}$ GeV is ruled out, leaving the $d=6$ island as the only genuinely new post-inflationary target.","The islands should be read as an upper bound on viable models, since the neglected topological-defect contribution to the axion relic density can only shrink them or make them disappear.","An axion signal at a mass excluded in standard cosmology would be an indirect signature of a self-induced early matter domination phase, a nonstandard cosmic history with no extra new physics.","The same combination of operator enumeration, Landau pole preselection, and Boltzmann cosmology could be applied to other long-lived coloured particle scenarios to map their cosmologically consistent parameter regions."],"forward_implications":["Haloscope experiments searching for axion masses around 10 µeV gain a motivated window, since the $d=6$ island corresponds to $f_a \\sim 10^{12}$ GeV.","A detection of an axion inside either island would point to a period of early matter domination and could constrain the effective operator scale $\\Lambda_{\\mathrm{EFT}}$.","The central band of the axion-photon coupling is stable across $f_a$, making the core predictions insensitive to the catalogue's parameter choices.","Models with domain wall number $N_{\\mathrm{DW}} = 1$, which avoid the domain wall problem, are a subset that yields generally larger couplings $|C_{a\\gamma}|$ and are particularly attractive search targets.","The publicly released catalogues allow the community to recompute the axion band under different assumptions, for instance including topological defect contributions."],"supporting_citations":[{"why":"Shows that higher-dimensional decay operators can self-consistently trigger a phase of early matter domination, which this paper extends to multiple heavy quarks.","marker":"[21]"},{"why":"Introduces the 'preferred model' selection criteria and the decay-operator dimension limit d≤5 for standard cosmology.","marker":"[17]"},{"why":"Derives the hadronic axion window using the same selection criteria and defines the original preferred hadronic models.","marker":"[18]"},{"why":"Provides the previous KSVZ model catalogue with multiple heavy quarks that this work extends to higher-dimensional operators.","marker":"[19]"},{"why":"Supplies the lattice QCD temperature-dependent axion mass used in the realignment calculation.","marker":"[30]"},{"why":"Provides the dark matter relic density bound and cosmological parameters used to set the upper limits on f_a.","marker":"[34]"},{"why":"Gives the tensor-to-scalar ratio bound that sets the Gibbons-Hawking temperature constraint on the post-inflationary scenario.","marker":"[70]"}],"fun_headline_variants":["New hadronic axion islands at 10^12 and 10^14 GeV","Axion model islands beyond standard mass window","d=6 and d=7 operators open new axion mass islands","Two new axion mass ranges made viable by heavy fermions","Hadronic axion models get new cosmological windows"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The post-inflationary interpretation requires that the Peccei-Quinn symmetry is unbroken during inflation, meaning $f_a$ must lie below the Gibbons-Hawking temperature $T_{\\mathrm{GH}} \\approx 6\\times 10^{12}$ GeV, a condition that the $d=7$ island at $f_a \\sim 10^{14}$ GeV violates.","fun_headline_variants_meta":{"raw":{"variants":["New hadronic axion islands at 10^12 and 10^14 GeV","Axion model islands beyond standard mass window","d=6 and d=7 operators open new axion mass islands","Two new axion mass ranges made viable by heavy fermions","Hadronic axion models get new cosmological windows"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000929,"raw_usage":{"total_tokens":3968,"prompt_tokens":925,"completion_tokens":3043,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":2957}},"tokens_in":541,"tokens_out":3043,"duration_ms":18458,"temperature":1.0,"reasoning_tokens":2957,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T05:09:40.563803+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the tensor-to-scalar ratio $r$ at the level of about 0.001, which would fix the inflationary Hubble scale at $H_I \\approx 6\\times 10^{12}$ GeV and the Gibbons-Hawking temperature at $T_{\\mathrm{GH}} \\approx 10^{12}$ GeV; with $T_{\\mathrm{GH}}$ inside the $d=6$ island's range ($f_a$ up to $2\\times 10^{12}$ GeV), the post-inflationary interpretation for that island would be ruled out, eliminating the paper's main new prediction.","supporting_citations":[],"review_version":1}