{"id":"6d39b8c3-59ee-4dcd-901e-e123cbd0cc2e","arxiv_id":"1908.03559","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Twin electrons and positrons, frozen in via a massive twin photon, can be the dark matter in mirror twin Higgs models with asymmetric reheating, with the required kinetic mixing matching loop-level expectations.","lead":"The paper proposes that dark matter could be made of mirror twin particles that are produced slowly after a period of cosmic dilution, rather than being left over in abundance. It shows the tiny coupling needed is the size physicists would naturally expect from loop effects in the mirror twin Higgs theory.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Residual twin e+e- from the pre-reheating plasma may swamp the tiny freeze-in yield; the required dilution is not quantified.","rationale":"The reader's weakest_assumption focuses on the uncomputed IR kinetic-mixing estimate in Appendix A, and that is indeed a genuine weakness in the paper's motivation. However, the model has explicit escape hatches for that issue: if the IR value of ϵ is larger than the required value, one can choose a larger m_γ' (up to the quoted 2m_π0 limit), and if it is smaller, UV contributions can supply the needed ϵ. By contrast, the freeze-in mechanism itself depends on the twin sector being essentially empty at T ~ 1 GeV. The standard MTH cosmology before asymmetric reheating contains a thermal twin e+e- plasma, and the paper's own choice m_γ' > m_e' removes the usual annihilation channel to twin photons. The resulting thermal twin e± relic must be diluted by a huge factor to fall below the very small freeze-in yield, but the paper does not quantify whether the cited asymmetric reheating models achieve this. This is a more direct threat to the central claim that the observed DM is the freeze-twin abundance. The check I propose is concrete and computationally straightforward: run the twin-sector Boltzmann equations with the benchmark dilution histories to see whether the residual e± yield is below Y_fi. If it is, the initial condition is justified and the freeze-in calculation stands; if not, the model overproduces DM and needs an additional depletion mechanism or a different m_γ'/m_e' hierarchy. This does not change the overall verdict from CONDITIONAL, because the model could still be viable with a sufficiently strong dilution, but it sharpens the condition that must hold and identifies a missing quantitative step in the paper.","tokens_in":13642,"tokens_out":36329,"duration_ms":417244,"concrete_test":"Use the asymmetric reheating benchmark of [14] (softly-broken Z2 scalar, f/v = 4) and integrate the twin-sector Boltzmann equations for e±, including annihilation to γ'γ' (kinematically suppressed below T ~ m_γ') and to ν'ν' via the twin Z, starting from full thermal equilibrium at T_twin ~ 1 GeV. Track the scale-factor dilution from the modulus decay and compute the twin e± yield at T_SM ~ 1 MeV. Compare with the freeze-in yield Y_fi ≈ 2×10^-7 for m_e' = 2 MeV. If Y_res > Y_fi, the central claim fails for that benchmark; if Y_res < Y_fi, the initial condition is justified. Repeat for the right-handed-neutrino asymmetric reheating mechanism of [42].","verdict_should_be":"UNCHANGED","load_bearing_attack":"The freeze-in yield is tiny (Y_e' ≈ 2×10^-7 for m_e' ~ 2 MeV), but the MTH before asymmetric reheating necessarily contained a thermal twin e+e- plasma, since the two sectors remain in equilibrium down to T ~ 4 GeV via the Higgs portal. Because the model requires m_γ' > m_e', the standard annihilation e+e- → γ'γ' is kinematically forbidden for nonrelativistic twin electrons and positrons; only highly suppressed annihilations to twin neutrinos through the heavy twin Z remain. The twin e± therefore freeze out with a large relic yield within the twin sector. Asymmetric reheating must dilute this yield by orders of magnitude beyond what is needed to satisfy ΔN_eff (T_twin/T_SM ~ 0.5 gives s_twin/s_SM ~ 0.2). The paper states 'we take the absence of twin energy density as an initial condition' (Sec. II) and cites [14,42], but gives no estimate of the residual twin e± number density after dilution for m_γ' > m_e'. If the residual yield exceeds the freeze-in yield, the dark matter is not freeze-twin and may overclose by orders of magnitude. This is more load-bearing than the Appendix A loop estimate, because that estimate has explicit escape hatches (larger m_γ' or UV ϵ), whereas a non-negligible thermal twin e± relic cannot be removed within the model. A quantitative check of the dilution factor in the cited asymmetric-reheating benchmarks is needed.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a freeze-in dark matter scenario in the Mirror Twin Higgs (MTH) framework with asymmetric reheating. A Stueckelberg mass is introduced for the twin photon, and the freeze-in production of twin electrons and positrons proceeds through the kinetic mixing between the SM and twin hypercharge/photons. The authors compute the yield analytically in the narrow-width and Maxwell-Boltzmann approximations, validate their numerical implementation against existing freeze-in computations, and derive the value of the kinetic mixing epsilon required to match the observed dark matter abundance. They find epsilon roughly in the range 10^-13 to 10^-10, which they argue matches the order expected from infrared loop contributions in the MTH. They also discuss constraints from supernova cooling and the possibility of self-interacting dark matter.","tokens_in":13892,"tokens_out":11868,"duration_ms":126696,"significance":"If the central assumption about initial conditions can be justified, this paper provides a minimal and well-motivated dark matter candidate in the MTH: the only new parameter is the twin photon mass, and both the feeble coupling and the negligible initial dark-matter abundance are motivated by physics orthogonal to dark matter. The freeze-in calculation is carefully done, with explicit validation against Refs. [62,64], and the resulting relation between m_gamma' and epsilon is a falsifiable prediction testable by supernova cooling and self-interaction observations. The authors are also transparent about the heuristic nature of the infrared loop estimate in Appendix A. The main weakness is the unquantified assumption about the absence of twin charged states after asymmetric reheating, which is load-bearing for the claim that the dark matter is produced by freeze-in.","major_comments":[{"comment":"The paper assumes that at T~1 GeV the twin sector is empty, stating \"we take the absence of twin energy density as an initial condition\" (Sec. II). However, in the MTH the twin sector is in thermal equilibrium with the SM down to T~4 GeV via the Higgs portal [14], so a thermal population of twin electrons and positrons exists before asymmetric reheating. For the parameter space of interest, m_gamma' > m_e', so the annihilation e+ e- -> gamma' gamma' is kinematically forbidden for non-relativistic pairs, and annihilation to twin neutrinos through the heavy twin Z is negligible. The twin e+/- therefore freeze out with a relic yield that is not diluted by the factor needed to satisfy Delta N_eff. Since the freeze-in yield is only Y_e' ~ 2x10^-7 (Sec. IV), the paper needs a quantitative check that in the asymmetric-reheating models of [14,42] the residual twin e+/- yield after dilution is below the freeze-in yield. Without this, the final dark-matter abundance could be dominated by the thermal relic rather than by freeze-in, and the central claim of the paper fails.","section":"II (initial conditions) and IV (freeze-in yield)"},{"comment":"The abstract and introduction state that the required kinetic mixing \"is of the loop-suppressed order expected from infrared contributions in the MTH.\" This expectation rests on the vanishing of lower-loop diagrams and on dimensional estimates of a four-loop diagram that has not been computed; the authors themselves note in Appendix A that they \"know no argument that kinetic mixing of this order is not generated.\" Because the match between the required epsilon and the expected IR range is a key motivation for the model, the paper should either present an explicit calculation (even a rough one) of the leading IR contribution or modify the abstract and introduction to clearly present this as a heuristic consistency check rather than a quantitative prediction.","section":"Appendix A and abstract"}],"minor_comments":[{"comment":"The caption of Fig. 1 should specify the values of f/v for each contour and explain the dashed segments and the shaded region; the plot is currently only described in the text in a piecemeal way.","section":"Fig. 1"},{"comment":"The neglect of the dT g_*s term is mentioned only briefly; a parenthetical estimate of its effect for m_gamma' near the QCD scale would help the reader assess the claimed 50% accuracy.","section":"Eq. (5)"},{"comment":"The statement that rho_twin ~ 0 is used as an initial condition would be clearer if it explicitly stated that the number densities of all twin species, including non-relativistic electrons and positrons, are assumed to be negligible after asymmetric reheating.","section":"Sec. II"},{"comment":"The paper does not discuss the lifetime of the massive twin photon and its subsequent decays into SM fermions; since the twin photon can decay through the same kinetic mixing, a comment on whether late-time decays of twin photons produced by dark-matter annihilations are compatible with indirect-detection constraints would be useful.","section":"Sec. IV"}],"recommendation":"major_revision","confidential_remarks":"The initial-condition issue in Section II is the key blocker in my assessment. I recommend asking for a concrete asymmetric-reheating benchmark where the residual twin e+/- yield is explicitly computed and shown to be small compared to the freeze-in yield. The loop estimate is more of a presentation issue; the paper is otherwise technically sound and clearly written."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here’s the short version: the freeze-in yield calculation is solid and the paper is honest about its approximations, but the residual thermal twin e+e- from the pre-reheating epoch is a bigger problem than the uncomputed loop estimate in Appendix A.\n\nWhat’s new: combining freeze-in with asymmetric reheating in the MTH, and showing the required kinetic mixing (ϵ ~ 10^-13–10^-10) is in the range one would guess from IR loop contributions. That’s a nice consistency observation, not a prediction; the authors say plainly they know no argument against generating it at that order. The calculation itself is standard, with Maxwell-Boltzmann statistics, narrow-width, and deliberate 50% accuracy. The validation against [62,64] looks real. The parameter space with supernova and self-interaction constraints is well presented.\n\nThe soft spots, in order of size. First, the residual twin e+e- issue. Before asymmetric reheating, the MTH has both sectors in equilibrium down to ~4 GeV. So a thermal population of twin e± existed. Because the model requires m_γ' > m_e', the usual annihilation e+e- → γ'γ' is kinematically blocked at low velocities, while annihilation to twin neutrinos is far too slow. The twin e± freeze out with a large yield per twin entropy. Asymmetric reheating dilutes the twin entropy relative to the SM, but even with T_twin/T_SM ~ 0.1 (s_twin/s_SM ~ 1e-3), the surviving Y_e' ~ 1e-4—about a thousand times the freeze-in yield from Eq. (9). The paper’s “we take the absence of twin energy density as an initial condition” is doing a lot of work; for radiation that is a familiar move, but for non-annihilating matter it requires the reheating mechanism to truly empty the twin sector, which none of the cited benchmarks demonstrably does. If Twinflation or something similar is the answer, the paper should say so and show how e± stay out.\n\nSecond, Appendix A: the IR mixing estimate is an order-of-magnitude guess from loop counting, with no diagram computed. The authors are upfront, and it is fine as motivation. But it should be framed as a benchmark, not a prediction.\n\nMinor: the Stueckelberg mass UV completion caveats are acknowledged, and I don’t count that against them.\n\nNet: a good idea, cleanly executed on the freeze-in side, but the cosmological initial condition is not innocent. It deserves peer review, but the referee should demand a quantitative dilution check for the thermal twin e± in the cited asymmetric-reheating scenarios.","headline":"A clean freeze-in calculation for twin dark matter sits on top of an unsupported initial condition: the pre-reheating twin e+e- plasma must be erased, and that is the real soft spot.","tokens_in":14454,"tokens_out":7602,"would_cite":true,"duration_ms":87318,"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":"Twin electrons frozen in through a massive twin photon can account for the observed dark matter abundance.","keywords":["mirror twin Higgs","freeze-in","dark matter","kinetic mixing","Stueckelberg mass","twin photon","asymmetric reheating","naturalness"],"falsifier":"Compute the four- and five-loop diagrams that could generate kinetic mixing in the low-energy mirror twin Higgs theory; if a complete calculation puts $\\epsilon$ outside the $10^{-13}$ to $10^{-10}$ window while $m_{\\gamma'}$ stays below a few hundred MeV, freeze-twin dark matter cannot produce the full observed abundance. Equally, a supernova neutrino measurement that excludes the required $m_{\\gamma'}$ versus $\\epsilon$ curve would falsify the scenario.","tokens_in":1675,"feed_emoji":"🌌","tokens_out":1573,"duration_ms":76006,"temperature":0.7,"pith_summary":"The paper tries to show that dark matter can be explained within the mirror twin Higgs model without adding any new tuning. After asymmetric reheating has diluted the twin sector to satisfy cosmological bounds, the twin sector is empty and cold enough that twin electrons and positrons can be produced slowly by freeze-in through a tiny kinetic mixing between the Standard Model photon and a Stueckelberg-massive twin photon. The mixing needed to reproduce the observed dark matter abundance, $\\epsilon \\sim 10^{-13}$ to $10^{-10}$, is the same size that the authors expect from infrared loop contributions within the model itself. If correct, this would turn the model's biggest cosmological obstacle into the setup for a natural dark matter candidate.","feed_headline":"Freeze-in through a massive twin photon can make all dark matter","feed_subtitle":"Asymmetric reheating empties the twin sector, then twin electrons and positrons freeze in to the observed abundance.","key_machinery":"The central object is the kinetic-mixing portal, $\\frac{\\epsilon}{2}F_{\\mu\\nu}F'^{\\mu\\nu}$, together with a Stueckelberg mass for twin hypercharge, $\\frac{1}{2}m_{\\gamma'}^2 A'_\\mu A'^\\mu$, which gives every SM fermion of electric charge $Q$ an effective twin charge $\\epsilon Q$. The mechanism is resonant freeze-in through $f\\bar f \\to \\gamma' \\to e'\\bar e'$, with the narrow-width approximation reducing the Boltzmann yield to a single integral over the SM bath temperature; production peaks near $T \\sim m_{\\gamma'}$, and the final abundance is inversely tied to the twin photon mass through the decay width $\\Gamma_{\\gamma'}$.","core_discovery":"The central claim is that the observed dark matter can be composed entirely of twin electrons and positrons, produced through freeze-in after asymmetric reheating has depleted the twin sector. Production is dominated by on-shell twin photons: SM fermion-antifermion pairs annihilate to $\\gamma'$, which decays to $e'\\bar e'$, and the yield is set by one new parameter, the Stueckelberg mass $m_{\\gamma'}$. The kinetic mixing required to match the observed abundance, obtained by inverting the relic-density condition, falls in the range $\\epsilon \\sim 10^{-13}$ to $10^{-10}$, which the authors argue is exactly the order expected from irreducible infrared loop contributions in the mirror twin Higgs. Thus the model is, in principle, an effectively parameter-free extension of the MTH with asymmetric reheating: the twin photon mass is the only free input, and the abundance then fixes the mixing at a naturally predicted size.","pith_inferences":["A direct calculation of the four- and five-loop diagrams that could generate kinetic mixing would settle whether the coincidence is real; if the mixing comes out at the bottom of the natural range, the model predicts a lighter twin photon, and if it comes out larger, the twin photon mass is pushed upward.","The same freeze-in logic extends to heavier twin photons, but then the yield is set during the asymmetric reheating epoch itself, so a measurement of the twin-sector dilution history (for instance through extra radiation in the cosmic microwave background) would be needed to make a precise prediction.","The mechanism turns the twin sector's emptiness into an asset: it predicts a dark matter candidate that interacts with our sector only through a tiny photon-like portal, giving stellar cooling and light-mediator searches a concrete target tied to the naturalness scale.","If hidden cancellations suppress the infrared loops, the model would likely require ultraviolet contributions to set $\\epsilon$, weakening but not destroying its parameter-free appeal; this is an explicit open point the paper identifies."],"forward_implications":["If freeze-twin dark matter is correct, the dark matter mass is set by $f/v$, the ratio that determines the twin spectrum, and collider measurements of Higgs couplings can therefore fix the dark matter mass.","Future supernova neutrino observations could probe the same $m_{\\gamma'}$ versus $\\epsilon$ curve that produces the relic abundance, since anomalous stellar cooling already constrains part of that plane.","For part of the allowed parameter space, twin-electron self-interactions fall in the range that has been suggested to address small-scale structure problems, and would be testable through cluster mergers and other astrophysical observations.","If the required $\\epsilon$ turns out larger than the infrared expectation, the scenario points to a heavier twin photon whose abundance depends on the details of asymmetric reheating; if smaller, ultraviolet contributions can still supply the needed mixing.","The scenario keeps the $Z_2$ symmetry that protects the Higgs mass, so a UV completion need not introduce hard mirror-symmetry breaking."],"supporting_citations":[{"why":"Introduces the mirror twin Higgs model whose $Z_2$ symmetry and Higgs-portal structure this work builds on.","marker":"[1]"},{"why":"Provides the decoupling and asymmetric reheating calculation that leaves the twin sector dilute, the starting condition for freeze-in.","marker":"[14]"},{"why":"Establishes the experimental bound $f \\gtrsim 3v$ and the soft-breaking framework used for the benchmark values.","marker":"[17]"},{"why":"Offers a concrete asymmetric reheating mechanism through right-handed neutrino decays, supporting the assumed dilute initial state.","marker":"[42]"},{"why":"Supplies the freeze-in yield formalism and the inverse-decay expression with which Eq. (9) agrees.","marker":"[47]"},{"why":"Used to validate the numerical freeze-in implementation through portal processes.","marker":"[62]"},{"why":"Gives one of the supernova cooling bounds that constrain the required kinetic mixing.","marker":"[69]"},{"why":"Provides the second, slightly different supernova bound included in the constraint band.","marker":"[70]"}],"fun_headline_variants":["Massive twin photon freeze-in yields all dark matter","Twin electrons and positrons freeze in as dark matter","One parameter sets twin electron dark matter abundance","Loop-suppressed mixing sets twin dark matter abundance","Asymmetric reheating enables twin freeze-in dark matter"],"cache_read_input_tokens":16512,"weakest_assumption_plain":"The whole scenario hinges on the expectation that loop effects in the mirror twin Higgs generate a kinetic mixing of order $\\epsilon \\sim 10^{-13}$ to $10^{-10}$; the authors themselves state that they know no argument that such a mixing is not generated, and hidden cancellations or larger ultraviolet contributions would break the coincidence.","fun_headline_variants_meta":{"raw":{"variants":["Massive twin photon freeze-in yields all dark matter","Twin electrons and positrons freeze in as dark matter","One parameter sets twin electron dark matter abundance","Loop-suppressed mixing sets twin dark matter abundance","Asymmetric reheating enables twin freeze-in dark matter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001679,"raw_usage":{"total_tokens":6614,"prompt_tokens":858,"completion_tokens":5756,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":474,"completion_tokens_details":{"reasoning_tokens":5682}},"tokens_in":474,"tokens_out":5756,"duration_ms":37751,"temperature":1.0,"reasoning_tokens":5682,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:09:40.602491+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the four- and five-loop diagrams that could generate kinetic mixing in the low-energy mirror twin Higgs theory; if a complete calculation puts $\\epsilon$ outside the $10^{-13}$ to $10^{-10}$ window while $m_{\\gamma'}$ stays below a few hundred MeV, freeze-twin dark matter cannot produce the full observed abundance. Equally, a supernova neutrino measurement that excludes the required $m_{\\gamma'}$ versus $\\epsilon$ curve would falsify the scenario.","supporting_citations":[{"cited_title":"Non-thermal right-handed sneutrino dark matter and the Omega_DM/Omega_b problem","cited_arxiv_id":"hep-ph/0701266","evidence_quote":"Supplies the freeze-in yield formalism and the inverse-decay expression with which Eq. (9) agrees."}],"review_version":1}