{"id":"79c9a239-87d1-4dd5-89c6-dada0d216d3a","arxiv_id":"2501.07725","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"In-flight annihilation of positrons produced by supernova feebly interacting particles gives the strongest astrophysical bounds on their electron couplings for masses of roughly 10-200 MeV, covering ALPs, sterile neutrinos, and dark photons.","lead":"The paper shows that positrons produced when heavy, weakly interacting particles from supernovae decay can annihilate while still moving fast, creating a gamma-ray glow that old telescopes already constrain. Using COMPTEL and EGRET data, it derives new upper limits on how strongly axion-like particles, sterile neutrinos, and dark photons couple to electrons, with the strongest gains at masses of roughly 10 to 200 MeV.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The IA signal is normalized to the model's own 511 keV line in the same ROI, so any bias in that anchor propagates directly into every headline coupling exclusion; a direct absolute IA calculation from the propagated positron spectra is required.","rationale":"The reader's weakest_assumption identifies the same issue that I consider most load-bearing. The IA signal is the entire observable of the paper; if its absolute normalization is biased, every exclusion limit shifts, and the claimed improvement over the 511 keV probe could shrink or disappear. The concern is grounded in the text's explicit statement that the IA normalization is fixed by the 511 keV emission in the region of interest. I do not find a more severe internal inconsistency: the production spectra and propagation setup are standard, and the multiple background treatments (fixed, scaled, power-law, no-background) are a strength rather than a weakness. However, the shared-normalization issue is real and currently untested. The proposed direct IA calculation would settle whether the anchor is reliable. Because the paper is otherwise internally consistent and transparent, the conditional verdict remains appropriate; no change to the reader's verdict is needed.","tokens_in":23218,"tokens_out":6719,"duration_ms":71058,"concrete_test":"For a benchmark case (e.g., 100 MeV ALP with gap = 2e-11 and gae = 1e-16), compute the IA gamma-ray flux in the ROI in two independent ways: (1) the paper's 511-anchored method; (2) a direct integration over the DRAGON2 propagated positron phase-space density of dN_gamma/dE_gamma = sigma_ann(E_pos) v n_e f_e(x, E_pos), using the same ISM electron density and the same energy-loss rates, normalized only by the injection spectrum and propagation. If the two absolute normalizations differ by more than ~30% at 20-60 MeV, where COMPTEL provides the constraints, the headline coupling limits must be revised; if they agree, the shared-normalization concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that in-flight annihilation (IA) gives the strongest FIP bounds rests on the absolute normalization of the IA flux. In Section III the authors state that the IA continuum is computed by first evaluating the integrated 511 keV line emission in the ROI and then fixing the IA normalization via the para-positronium-to-IA ratio at 511 keV. The 511 keV anchor is not an external observable: it is the authors' own prediction for the same FIP model, from the same DRAGON2 propagation and the same thermalization/positronium assumptions. Consequently, any systematic error in the propagated positron number, in the energy-loss rate that sets the thermalization time, or in the positronium fraction enters both the 511 keV line and the IA flux in the same direction and is not cancelled by comparing IA to COMPTEL data. Since the headline limits (e.g., gae < 1e-17 for ma = 10-100 MeV ALPs, |U_mu4|^2 < 4e-14, epsilon > 7e-14 for DPs with fixed background) scale as the square root of the flux, a factor of 2-3 error in the 511 anchor changes the coupling exclusions by a corresponding factor, and a larger error would weaken the claimed improvement over the 511 keV probe. The paper never presents a direct first-principles computation of the IA flux from the propagated positron distribution, so this normalization is untested.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper computes the diffuse gamma-ray signal from in-flight annihilation (IA) of positrons produced by decays of feebly interacting particles (FIPs) emitted by Galactic core-collapse supernovae, and uses it to constrain ALPs coupled to electrons (and separately to nucleons plus electrons), sterile neutrinos mixed with muon or tau neutrinos, and dark photons kinetically mixed with the photon. FIP production spectra are computed for the Garching SFHo-s18.8 SN model; the injected e± are propagated with a customized DRAGON2 setup; and the IA continuum is normalized, following the Beacom-Yuksel approach, to the model's own integrated 511 keV line emission in the region |l| < 60°, |b| < 10°. Two-sigma coupling limits are derived from COMPTEL data under four background treatments (none, power-law, reference model fixed, reference model with free normalization), with propagation uncertainties bracketed by extreme halo-height and Alfven-velocity scenarios. The headline exclusions are gae ≲ 10^-17 for 10-100 MeV nucleon-plus-electron-coupled ALPs at gap = 2×10^-11, |U_μ4|^2 ≲ 4×10^-14 and |U_τ4|^2 ≲ 6×10^-14 for sterile neutrinos, and ε ≳ 7×10^-14 for dark photons (fixed background), claimed to improve previous astrophysical bounds by one to two orders of magnitude.","tokens_in":23537,"tokens_out":27857,"duration_ms":257642,"significance":"If the absolute IA normalization holds, the paper demonstrates a genuinely more powerful astrophysical probe of FIP-electron couplings than the 511 keV line or the previously used IC and bremsstrahlung channels: the IA continuum peaks at tens of MeV where the diffuse gamma-ray background is low, and the three FIP models are computed from explicit production spectra rather than a generic parametrization. Strengths include the use of a modern SN model, the full DRAGON2 propagation treatment, a transparent menu of background and propagation scenarios, conservative choices (coldest SN profile, no-background limits), and falsifiable MeV-range predictions that next-generation telescopes can test. The main correctness risk is that the IA normalization is anchored to the same pipeline that predicts the 511 keV line, so the absolute exclusions inherit unquantified systematics; the relative claim that IA is the strongest positron channel is robust to this risk, while the absolute numbers are not. The authors also deserve credit for displaying the full spread between the most and least conservative analyses.","major_comments":[{"comment":"The absolute normalization of the IA signal is fixed by the model's own integrated 511 keV line emission in the same region of interest, since the text states that 'the normalization is fixed by the 511 keV emission in that region of interest' and that 511 keV emission is itself a prediction of the same pipeline (same FIP injection spectra, same DRAGON2 propagation, same thermalization and positronium assumptions). A systematic error in the propagated positron yield, thermalization fraction, or positronium fraction therefore enters the IA spectrum and the 511 keV anchor coherently, and comparing the IA signal to COMPTEL data does not cancel this common mode. The paper does not present a direct IA calculation from the steady-state positron distribution that would validate the anchored normalization, nor does it propagate anchor systematics into the limits of Fig. 9; the IA method is only described by reference to Eq. (2) of the companion work [24]. Since the headline couplings scale with a fractional power of the IA flux (flux ∝ gae^2 and ∝ |U_alpha4|^4 in the cases shown in Fig. 9), an unquantified factor-2-3 anchor error shifts the quoted exclusions by factors of order 1.2-1.7. I request either (a) a direct first-principles computation of the IA flux from the propagated positron spectra that reproduces the anchored normalization, or (b) a systematic band associated with the anchor (positronium fraction, thermalization time, free-electron density) added to the limits.","section":"III (gamma-ray signals section)"}],"minor_comments":[{"comment":"The statement that bounds are derived 'using data from COMPTEL and EGRET' is inaccurate because Section III explains that EGRET data are used only for illustrative purposes and that all constraints are derived from COMPTEL; please rephrase and fix the typo 'strenghten' in the same paragraph.","section":"Abstract"},{"comment":"The sentence following Eq. (6) reads 'where where E−,+ and f−,+ are the electron and positron energies and distribution functions'; the duplicated 'where' should be removed, and the positron distribution function f+ should be defined explicitly since it is not introduced before this equation.","section":"II (after Eq. (6))"},{"comment":"The decay channels producing the e± pairs (presumably ν4 → νℓ e+e−) and the positron injection spectra used in Eq. (21) are not described; since the sterile neutrino limits are headline results, the decay mode, branching ratios, and how the three-body final state is mapped onto the Ea = xi Epos prescription of Eq. (21) should be stated in the text rather than referenced only to Refs. [15, 24].","section":"II (sterile neutrinos)"},{"comment":"The headline numbers (e.g., ε ≳ 7×10^-14 for DPs, gae ≲ 10^-17 for ALPs) correspond to the analysis with the reference fixed background, while the no-background analysis gives limits up to an order of magnitude weaker; the abstract should state this dependence explicitly so that the claimed 'one to two orders of magnitude' improvement is not read as background-independent.","section":"Abstract and Section IV"},{"comment":"The text says the propagation accounts for 'diffusion-reacceleration-advection-loss effects,' but Eq. (24) displays no advection term; please clarify whether advection is neglected for sub-GeV positrons and explain the justification.","section":"III (Eq. (24) discussion)"},{"comment":"The e± injection tables are stated to be available 'upon request'; depositing them in a permanent public repository would improve reproducibility, particularly because the IA normalization is otherwise recoverable only from the companion work [24].","section":"III (propagation setup)"},{"comment":"The caption contains a typo: 'This underscores the important of IA at \"higher\" masses' should read 'the importance of IA'.","section":"Fig. 10 caption"},{"comment":"The FIP emission-time interval written as 'i.e. 1 − 10 s' should be '1–10 s' (en dash), since '1 − 10 s' is ambiguous.","section":"II (Eq. (21) discussion)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a close follow-up to the authors' Ref. [24], where the IA method and sterile-neutrino bounds already appeared in a generic-parametrization form; the genuinely new content here is the model-by-model SN production spectra and the resulting ALP and DP limits. The main technical caveat is the IA normalization's dependence on the model's own 511 keV prediction; if the authors supply a direct cross-check or an explicit anchor-systematics band, the paper would be suitable for a strong particle-astrophysics journal. The incremental advance over Ref. [24] should be articulated more sharply in the introduction. There is substantial self-citation of the companion papers, which is understandable given that the same authors built the pipeline, but a sentence situating the present work relative to Ref. [24] would help the reader."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Pedro, this is a solid phenomenological paper that does what it says: takes the in-flight annihilation (IA) mechanism the authors introduced for sterile neutrinos and applies it with full SN production spectra to ALPs (electron- and nucleon-coupled), dark photons, and sterile neutrinos. The production spectra for each model are genuinely new and carefully derived, and the comparison to COMPTEL is done with several background treatments and propagation uncertainties. Credit where due: the pipeline is complex and mostly internally consistent, and the authors are honest that the dark photon bounds are complementary to cosmological constraints, even if the abstract overstates it.\n\nThe soft spot is the one flagged in the stress test. The IA normalization is fixed by the model's own integrated 511 keV emission in the same ROI, not by an observed 511 keV flux. So any systematic error in the propagated positron population, thermalization fraction, or positronium ratio enters both the 511 keV anchor and the IA spectrum in the same direction. The limits then scale as the square root of the flux, so a factor 2-3 anchor error moves the exclusions by a corresponding factor. That is not fatal, and it is not circular in the strict sense — the final limits compare signal plus background to COMPTEL data, and no fit parameter reduces to the anchor. But it means the absolute normalization of the headline IA limits is untested in this paper. A direct first-principles IA calculation from the propagated positron distribution would settle it. The companion work [24] apparently contains details, but I would want that shown here or at least a sensitivity test where the 511 anchor is varied by a factor 2.\n\nOther points: the code and injection tables are only available on request, which makes independent reproduction harder. The 'strongest bounds' claim for heavy ALPs and sterile neutrinos in the 10-200 MeV range is plausible and well supported by the comparison with the 511 keV line limits shown in Fig. 9. The EGRET data are appropriately used only illustratively.\n\nWho is this for? Anyone working on MeV-scale FIP constraints or future MeV gamma-ray missions. It deserves a serious referee; the referee should push for releasing the pipeline and for a direct IA absolute normalization check, but the central argument holds up as a conditional result. I would send it to review.","headline":"Extends the in-flight annihilation probe to ALPs, dark photons, and sterile neutrinos with full supernova production spectra, but the IA normalization anchored to the model's own 511 keV prediction leaves the headline limits hostage to that anchor.","tokens_in":24102,"tokens_out":2576,"would_cite":true,"duration_ms":25785,"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":"In-flight annihilation of supernova positrons gives the strongest astrophysical bounds on the electron couplings of heavy axion-like particles, sterile neutrinos, and dark photons, improving previous limits by one to two orders of…","keywords":["in-flight positron annihilation","feebly interacting particles","axion-like particles","sterile neutrinos","dark photons","supernova positrons","511 keV line","gamma-ray bounds"],"falsifier":"Measure the diffuse gamma-ray spectrum in the $|l|<60^\\circ$, $|b|<10^\\circ$ region with sensitivity in the few-MeV to 100 MeV band: a detected bump matching the IA shape at the level predicted for couplings just above the quoted limits would falsify the exclusions, while an independent determination of the 511 keV disk flux a factor of two below the adopted anchor would rescale every limit upward by the same factor.","tokens_in":23016,"feed_emoji":"⚛️","tokens_out":8789,"duration_ms":76624,"temperature":0.7,"pith_summary":"Core-collapse supernovae can emit feebly interacting particles (FIPs) heavy enough to decay into electron-positron pairs, and this paper argues that the sharpest trace of those positrons is the gamma-ray continuum they produce while annihilating in flight, not the classic 511 keV line. The paper computes this in-flight annihilation (IA) signal for axion-like particles, sterile neutrinos, and dark photons, propagates the injected positrons through the Galaxy, and compares the predicted gamma rays with COMPTEL and EGRET measurements of the Galactic plane. Its central claim is that the IA signal sets the most restrictive astrophysical limits on the electron couplings of these particles, excluding couplings one to two orders of magnitude beyond previous bounds depending on the model. A sympathetic reader would care because those couplings are otherwise very hard to probe: the IA gamma-ray bump sits at tens of MeV, where the astrophysical background is low and the annihilation cross section is large.","feed_headline":"Supernova positrons tighten new-particle limits 100-fold","feed_subtitle":"In-flight annihilation beats the 511 keV line as a probe of axions, sterile neutrinos, and dark photons.","key_machinery":"The load-bearing object is the in-flight annihilation spectrum of relativistic positrons on interstellar electrons. Its energy dependence is set by the injected positron spectrum from FIP decay, the energy losses during Galactic propagation, and the Dirac annihilation cross section; its normalization is anchored by the model's own 511 keV line emission in the same sky region, following a strategy in which the continuum is obtained from the measured para-positronium to IA ratio at 511 keV. This anchoring means the IA signal is not an independent flux: it inherits whatever errors affect the propagated positron number, the thermalization fraction, and the positronium fraction, and it scales linearly with those quantities.","core_discovery":"The paper's claim, stated on its own terms, is that in-flight positron annihilation from FIP decays in Galactic supernovae is the dominant secondary gamma-ray signal and the best available astrophysical probe of FIP-electron couplings. Normalizing the continuum to the 511 keV line in the same region of interest, the authors derive 2-$\\sigma$ exclusion limits: for nucleon-coupled ALPs with $g_{ap}=2\\times10^{-11}$, electron couplings $g_{ae}\\gtrsim 10^{-17}$ are excluded for $10\\,\\mathrm{MeV}\\lesssim m_a\\lesssim100\\,\\mathrm{MeV}$ (with resonant structure around 135 and 147 MeV); for sterile neutrinos mixed with muon or tau neutrinos, $|U_{\\mu 4}|^2\\gtrsim4\\times10^{-14}$ and $|U_{\\tau 4}|^2\\gtrsim6\\times10^{-14}$ are excluded for masses between 10 and 150 MeV; and for dark photons, $\\epsilon\\gtrsim7\\times10^{-14}$ is excluded with a fixed background model, or $\\epsilon\\gtrsim6\\times10^{-13}$ without any background, for masses between 5 and 50 MeV. These bounds are claimed to strengthen existing limits by one to two orders of magnitude.","pith_inferences":["A future MeV gamma-ray telescope with better angular resolution and lower backgrounds should sharpen these limits substantially, and could convert the current null into a detection if the true coupling lies just below the COMPTEL bound.","Because the IA normalization is inherited from the 511 keV line, improving the modeling of the Galactic disk's 511 keV emission would directly strengthen all three sets of limits without any new gamma-ray data.","The same IA method applies to any steady Galactic source of energetic positrons, including dark-matter decay or neutron-star merger remnants, so the framework is broader than supernova-produced FIPs.","The spectral shape difference (dark-photon IA peaking near 20-30 MeV, axion IA near 60 MeV) gives a potential diagnostic for identifying which particle produces a future signal."],"forward_implications":["If the central claim holds, the IA gamma-ray continuum replaces the 511 keV line as the most constraining astrophysical observable for electron couplings of MeV-scale FIPs produced in supernovae.","Nucleon-coupled ALPs in the 10-100 MeV mass range with $g_{ae}$ above about $10^{-17}$ would be excluded (for $g_{ap}=2\\times10^{-11}$), with the limits relaxing near the 135 MeV photon-decay resonance and tightening near 147 MeV.","Sterile neutrinos mixed with muon or tau neutrinos would be excluded for $|U_{\\alpha 4}|^2$ above roughly $4\\times10^{-14}$ (muon) or $6\\times10^{-14}$ (tau) across 10-150 MeV.","Dark photons in the 5-50 MeV range would be excluded for kinetic mixing $\\epsilon$ above about $7\\times10^{-14}$ with a fixed background, complementing cosmological and SN 1987A constraints.","In every background model considered, the COMPTEL data require no IA signal on top of the background, so the bounds are purely exclusion limits rather than evidence for new physics."],"supporting_citations":[{"why":"Companion work that introduced IA as a FIP probe and supplies the 511 keV-anchored normalization used here.","marker":"[24]"},{"why":"Provides the core-collapse supernova model whose temperature and density profiles set all FIP production fluxes.","marker":"[38]"},{"why":"Gives the electron-ion bremsstrahlung ALP production rate used for electron-coupled ALPs.","marker":"[56]"},{"why":"Gives the electron-positron fusion ALP production rate that dominates at higher ALP masses.","marker":"[57]"},{"why":"Provides the nuclear ALP emission calculation (nucleon bremsstrahlung plus pion conversion) used for nucleon-coupled ALPs.","marker":"[41]"},{"why":"Sets dark-photon production via proton bremsstrahlung and the envelope radii and supernova type fractions used in decay transport.","marker":"[86]"},{"why":"Supplies the customized Galactic cosmic-ray propagation code that evolves the injected positrons to steady state.","marker":"[95]"},{"why":"Provides the IA-versus-511 keV strategy from which the continuum normalization is derived.","marker":"[115]"},{"why":"COMPTEL diffuse gamma-ray data in the region of interest; the data set from which the exclusion limits are extracted.","marker":"[108]"},{"why":"Reference Galactic background model for inverse-Compton and bremsstrahlung emission used in the fixed-background limits.","marker":"[118]"}],"fun_headline_variants":["In-flight positrons tighten FIP limits 100-fold","Supernova positron signal sharpens axion and dark photon bounds","Positron annihilation in flight yields strongest FIP-electron limits","In-flight annihilation trumps 511 keV line for new-particle probes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The IA signal's absolute size is fixed by the model's own 511 keV line emission in the same sky region, so any error in the propagated positron number, the thermalization fraction, or the positronium fraction shifts the IA spectrum and all derived limits by the same factor.","fun_headline_variants_meta":{"raw":{"variants":["In-flight positrons tighten FIP limits 100-fold","Supernova positron signal sharpens axion and dark photon bounds","Positron annihilation in flight yields strongest FIP-electron limits","In-flight annihilation trumps 511 keV line for new-particle probes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000234,"raw_usage":{"total_tokens":1502,"prompt_tokens":958,"completion_tokens":544,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":574,"completion_tokens_details":{"reasoning_tokens":472}},"tokens_in":574,"tokens_out":544,"duration_ms":5816,"temperature":1.0,"reasoning_tokens":472,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:37:10.679333+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the diffuse gamma-ray spectrum in the $|l|<60^\\circ$, $|b|<10^\\circ$ region with sensitivity in the few-MeV to 100 MeV band: a detected bump matching the IA shape at the level predicted for couplings just above the quoted limits would falsify the exclusions, while an independent determination of the 511 keV disk flux a factor of two below the adopted anchor would rescale every limit upward by the same factor.","supporting_citations":[],"review_version":1}