{"id":"bc6d969a-751b-400b-8f8c-ab7e70c9110b","arxiv_id":"1304.1840","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"New AMS positron data rule out direct e+e- and mu+mu- dark matter annihilation but leave ~1-3 TeV dark matter annihilating to intermediate states, and pulsars, viable under large diffusion-zone assumptions.","lead":"This paper re-tests whether annihilating dark matter or nearby pulsars can explain the rising cosmic-ray positron fraction measured by AMS. It finds that direct annihilation to electrons or muons is now excluded, while ~1-3 TeV dark matter annihilating to intermediate states, or pulsars, can still fit under favorable propagation assumptions.","discovery_kind":"extension","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The exclusion of direct e+e-/mu+mu- annihilation and the survival of XDM are established under different, unequally favorable assumptions (L=4 kpc + simple power law vs L=8 kpc + broken power law), so the paper's central dichotomy is not tested on equal footing.","rationale":"The reader's weakest_assumption correctly locates the soft spot: DM viability hinges on L=8 kpc and the broken power law. I agree in substance but sharpen it. The more damaging issue is not merely that XDM viability is conditional, but that the exclusion of the competing direct channels is never tested under the same favorable conditions. The paper is otherwise transparent: it states plainly (Sec. II and the Fig. 2 caption) that the models fail under the default assumptions, and it justifies the spectral break via local-source stochasticity (footnote 1). The classifier's 'Limitations: none found' is a miss — the authors do acknowledge the contingency. The pulsar case (Figs. 7-8) is less contingent, and the 10Be/9Be forward look is a genuine falsifiable handle. So the paper's honesty is not in question; what remains unresolved is whether the exclusion/viability asymmetry is an artifact of unequal fitting assumptions. That is a bounded, checkable question. The verdict stays CONDITIONAL; my concern reinforces rather than overturns the reader's.","tokens_in":18536,"tokens_out":3656,"duration_ms":84580,"concrete_test":"Using the authors' GALPROP v54 setup, rerun the direct e+e- (350 and 900 GeV) and mu+mu- (600 GeV and 1.5 TeV) fits at L=8 kpc with the broken power-law background of Fig. 6, refitting <sigma v> to the AMS positron fraction. Report chi2/dof to AMS and Fermi. If the direct channels remain above ~3, the exclusion stands and the dichotomy is sound; if they fall to O(1), the paper's central contrast between 'excluded' and 'viable' DM models collapses and the claim must be restated as propagation-dependent for all channels.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline claim (abstract:s2) is a dichotomy: direct e+e- and mu+mu- annihilation 'no longer appear capable' of fitting AMS, while XDM to intermediate states 'could still be responsible.' The two halves rest on fits made under different assumptions. Direct-channel fits appear only in Fig. 1, at L=4 kpc with a simple power-law electron background, giving chi2/dof of 9.3-15.3 to AMS. The XDM models that survive appear in Figs. 5-6, at L=8 kpc with a broken power law (-2.65 to -2.3 at 85 GeV), where chi2/dof drops to 0.8-1.3. The paper never shows direct e+e- or mu+mu- at L=8 kpc with the broken background. Because Fig. 3 demonstrates that raising L softens the positron-fraction slope — exactly the feature that disfavors hard direct channels — the exclusion may not survive the same treatment that rescues XDM. The summary then states the exclusion flatly ('are no longer capable'), which outruns the L=4-only evidence. This is not a request to re-prove a cited result; it is an internal-consistency gap in the paper's own comparison of competing models.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"Cholis and Hooper re-examine the origin of the rising cosmic-ray positron fraction using the first AMS-02 measurement (0.5–350 GeV). Using GALPROP v54 with a cylindrical diffusion zone of half-width L = 1–8 kpc, they fit annihilating dark matter models (direct e+e− and μ+μ−, and 'eXciting' dark matter annihilating to intermediate states φ that decay to e+e−, μ+μ−, π+π−) and a pulsar population (sum of all Milky Way pulsars plus the two nearby young pulsars Geminga and B0656+14) to the AMS positron fraction and to the Fermi/HESS lepton spectrum. They conclude that direct annihilation to e+e− or μ+μ− can no longer accommodate the AMS data, that ~1–3 TeV dark matter annihilating to intermediate states can, and that pulsars remain viable. They further argue that future AMS measurements of secondary-to-primary ratios, especially 10Be/9Be, could discriminate between the surviving scenarios.","tokens_in":18850,"tokens_out":3320,"duration_ms":73513,"significance":"The question is timely and the AMS-02 positron-fraction measurement was, at the time of writing, the most constraining datum available, so a careful re-assessment of both the dark matter and pulsar interpretations is useful to the community. The paper is transparent in places where transparency is unusual: it states explicitly in the Fig. 2 caption that, under the propagation and background assumptions used there, all the models shown are excluded, and it states in the text that none of the models shown up to that point reproduce the Fermi lepton spectrum. The propagation modelling with GALPROP, the scan over L, and the separate treatment of the two nearby pulsars are all competently done. The paper does not ship reproducible code or machine-checked derivations, and its fits are not parameter-free; the value of the work therefore rests on whether the comparison between competing source scenarios is made on equal footing, which is the substance of the comments below.","major_comments":[{"comment":"The headline dichotomy in abstract:s2 — direct e+e−/μ+μ− annihilation 'no longer appear capable' while annihilation to intermediate states 'could still be responsible' — is established under two different and unequally favourable sets of assumptions. The direct-channel fits appear only in Fig. 1, at L = 4 kpc with a simple power-law electron background, giving χ²/d.o.f. of 9.3–15.3 to AMS. The XDM models that survive appear in Figs. 5 and 6, at L = 8 kpc with a broken power law, where χ²/d.o.f. drops to 0.8–1.3. Fig. 3 shows explicitly that raising L softens the slope of the predicted positron fraction, which is precisely the feature that disfavours the hard direct channels. The manuscript never shows the direct e+e− or μ+μ− channels at L = 8 kpc with the broken background, so the exclusion of those channels is not tested under the conditions that rescue XDM. The authors should either add those fits or soften the abstract and summary to state that the exclusion holds for L ≲ 4 kpc with a simple power-law background.","section":"§II, Figs. 1, 5, 6; abstract:s2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The broken power-law electron background (dNe/dEe ∝ E^−2.65 below 85–100 GeV, hardening to −2.3 above) is introduced only after the text states that 'in none of the cases we have shown so far have we been able to find good agreement with the cosmic ray electron+positron spectrum as measured by Fermi.' Footnote 1 concedes that the −2.65 slope is 'generally considered to be steeper than what is anticipated from 1st order Fermi accelerated electrons (≲ −2.0).' The break energy is not derived, and no scan over break energy or post-break slope is presented, so the reader cannot tell how much of the improvement in χ² in Fig. 6 comes from the dark matter model and how much from the tuned background. A scan over break energy (say 50–200 GeV) and post-break slope, showing the region in which each surviving model still fits, would make the claim quantitative rather than demonstrative.","section":"§II, Fig. 6 and footnote 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The χ²/d.o.f. values quoted throughout the figure captions are not reproducible as stated: the fit energy range is never given, the number of AMS and Fermi data points used is never given, and it is not stated whether the low-energy AMS points (below ~10 GeV, where charge-sign-dependent solar modulation is significant) are included or excluded from the fits. The notation '15.3(10.6)' is also ambiguous as to which number corresponds to which data set in some captions. Since the central claims are quantitative comparisons of χ² between models, these details are load-bearing and should be tabulated in the text rather than left implicit in the captions.","section":"§II, Figs. 1, 2, 4, 5, 6 (captions)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The pulsar case exhibits the same unequal-background structure as the dark matter case, and the abstract's statement that pulsars 'remain capable' (abstract:s3) glosses over it. In Fig. 7, with a simple power-law background, the top panel gives χ²/d.o.f. of 6.9 to Fermi, which is not an acceptable fit; only Fig. 8, with the same broken power law used to rescue XDM, gives 0.37. In addition, the required pair-production efficiency is quoted as 16% of total pulsar energy for the diffuse population but 3–4% once Geminga and B0656+14 are included, a factor of four to five discrepancy that is presented without comment. The efficiency is effectively a free normalization fitted to the same AMS data the model is said to explain, and the manuscript should say so plainly and state what independent constraint, if any, bounds it.","section":"§III, Figs. 7 and 8","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The abstract quotes '~1-3 TeV' for the surviving dark matter mass range, while §IV quotes '~1.5-3 TeV' and the models actually shown in Fig. 6 are 1.6, 2.5 and 3.0 TeV. The abstract range should be brought into agreement with the body.","section":"Abstract vs §IV","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption states that 'for the given propagation/background assumptions all these models are excluded,' which is a stronger and more honest statement than the abstract's 'no longer appear to be capable.' The wording of the abstract and of §IV should be reconciled with this caption so that the reader is not left with two different verdicts.","section":"Fig. 2 caption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Typographical: 'bellow' should read 'below'.","section":"Footnote 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Typographical: 'botom' should read 'bottom'.","section":"Fig. 7 caption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The diffusion coefficient normalization is said to be chosen to fit the non-leptonic secondary-to-primary data, but the numerical values adopted are never given. A short table of the propagation parameters used for each L would let the reader reproduce Figs. 3–6.","section":"§II, paragraph on propagation","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The injection cutoff Ec is fixed at 600 GeV throughout, while the text states that 80–1000 GeV covers the range found in the literature. A sentence justifying the fixed value, or a brief demonstration that the fit is insensitive to it, would strengthen the pulsar case.","section":"§III, pulsar spectral parameters","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Ref. [76] (Yuan et al., arXiv:1304.1482) is a contemporaneous paper addressing similar issues. The manuscript should state more specifically what the present analysis adds relative to that work, beyond the acknowledgment that it appeared while this project was being completed.","section":"Acknowledgments and Ref. [76]","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision","confidential_remarks":"The paper is short and its headline is an exclusion claim, so the unequal treatment of direct-channel versus intermediate-state dark matter (and of the simple versus broken electron background for pulsars) is the point on which the paper stands or falls. The fix is not expensive: run the direct channels at L = 8 kpc with the broken background and report the result, and tabulate the fit ranges and d.o.f. If the direct channels remain excluded under those conditions, the abstract's flat wording is earned; if they do not, the abstract needs to be rewritten. The near-simultaneous appearance of Yuan et al. (1304.1482) also warrants a clearer statement of what is new here."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"The useful thing here is the timing. AMS's 2013 positron fraction landed and Cholis & Hooper turned around a careful GALPROP re-analysis of both the dark matter and pulsar explanations. The concrete new result is that direct annihilation to e+e- or mu+mu- no longer fits: the AMS rise is too soft for those hard injection spectra. That is a real update over the PAMELA-era literature and worth knowing. The pulsar section is also honest. Figs. 7-8 show the summed Galactic pulsar population can do the job with 10-20% of spin-down energy into pairs, and Geminga plus B0656+14 at 3-4% if you let the two nearby young ones dominate locally. Credit where due: the propagation modeling is standard but done properly, they scan L = 1-8 kpc rather than picking one value, and the broken power-law electron background is physically motivated rather than invented. Above ~100 GeV the local electron spectrum should be dominated by a few nearby sources, so a hardening is expected, and the footnote offering local-source overdensity as an alternative is the right kind of caveat. The 10Be/9Be forward look is a genuine falsifiable handle and the best part of the discussion. Now the soft spot, and the stress test has it right. The exclusion of direct channels is shown only at L = 4 kpc with a simple power law (Fig. 1, chi2/dof 9-15). The XDM models that survive are shown at L = 8 kpc with the broken background (Figs. 5-6, chi2/dof 0.8-1.3). Fig. 3 shows that raising L softens the positron-fraction slope, which is exactly the feature that hurts hard direct channels. The paper never shows direct e+e- or mu+mu- at L = 8 with the break. So the dichotomy in abstract:s2 is not tested on equal footing, and the summary's flat 'are no longer capable' outruns the evidence. The abstract's 'appear to be' is better calibrated than the conclusion. One more figure would settle it and a referee should ask for it. Second, smaller: the required cross sections are (6-23) x 10^-24, i.e. 100-1000x thermal. The paper says so and lists the usual boost mechanisms without committing to any, which is fine for phenomenology but means 'DM could still be responsible' is really 'DM could still be responsible if something else supplies the boost.' Third, minor: the break energy is 85 GeV in Fig. 6 and 100 GeV in Fig. 8. Probably different fits, but worth a sentence. Who is it for: anyone tracking the positron excess, and pulsar people who want a benchmark. It deserves referees.","headline":"A timely, competent re-fit of the AMS positron fraction; the exclusion of direct leptonic annihilation is real but is argued on a different footing than the XDM models that survive.","tokens_in":19463,"tokens_out":2462,"would_cite":true,"duration_ms":61692,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["95.85.Ry","95.35.+d","97.60.Gb"],"model":"selfhosted/deepseek-v4.1-flash","headline":"The new AMS positron data rule out the simplest dark matter explanations for the cosmic ray positron rise, but leave two families of models standing: heavy dark matter that annihilates through short-lived intermediate particles, and nearbyp","keywords":["cosmic ray positron fraction","AMS-02","dark matter annihilation","pulsars","GALPROP propagation","exciting dark matter","PAMELA","lepton spectrum"],"falsifier":"An AMS measurement of the $^{10}$Be/$^9$Be ratio that pins the diffusion zone half-width below about 4 kpc would remove the propagation regime in which the surviving dark matter models fit, leaving pulsars as the only viable source among those considered here.","tokens_in":19910,"feed_emoji":"🌌","tokens_out":2600,"duration_ms":66099,"temperature":0.7,"pith_summary":"Cosmic ray positrons are a puzzle because ordinary cosmic ray collisions should produce far fewer of them at high energy than PAMELA and now AMS observe. The paper asks what could inject that many energetic positrons into the local neighborhood of the Milky Way, and tests two candidates against the much sharper AMS positron fraction: annihilating dark matter, and pulsars. The answer is a narrowing rather than a verdict. Dark matter that annihilates straight into electron-positron or muon pairs no longer fits, but dark matter in the 1-3 TeV range that annihilates into short-lived intermediate particles which then decay to muons or pions can still be fitted, at the cost of a large annihilation rate and a thick diffusion zone. Pulsars remain equally viable, needing only a few percent of their spin-down energy converted into electron-positron pairs.","feed_headline":"AMS data kills the simplest dark matter positron models","feed_subtitle":"Only annihilation through short-lived intermediates, or nearby pulsars, still fits the rising positron fraction.","key_machinery":"The argument is carried by a joint fit: GALPROP propagation of electrons and positrons through a cylindrical diffusion zone of half-width $L$, combined with a parametrized source term (dark matter annihilation spectrum or pulsar injection spectrum) and a primary electron background, compared simultaneously to the AMS positron fraction and the Fermi/HESS total lepton spectrum. The diffusion half-width $L$ is the dial that matters most. Increasing $L$ from 2 to 8 kpc softens the predicted positron fraction and improves the dark matter fits dramatically, and the paper is explicit that $L$ is currently constrained only loosely by secondary-to-primary ratios. The second piece of machinery is the intermediate-state cascade, which converts a hard injection spectrum into a softer one and is what allows the heavier dark matter models to survive.","core_discovery":"Using the AMS measurement of the positron fraction between 0.5 and 350 GeV, the authors re-fit annihilating dark matter models within the GALPROP propagation code and find that the direct leptonic channels are now disfavored: annihilation to $e^+e^-$ or $e^+e^-$ gives poor fits to the AMS shape, and the fits to the Fermi and HESS total lepton spectrum are worse. What survives is 'exciting dark matter', in which a 1-3 TeV particle annihilates into a pair of light intermediate states that subsequently decay to combinations of $e^+e^-$, $e^+e^-$, and $e^+e^-$, softening the injected positron spectrum. With a diffusion zone half-width of $L = 8$ kpc and a broken power-law electron background that hardens from $E^{-2.65}$ to $E^{-2.3}$ near 85-100 GeV, the surviving models reach $1.6$ TeV with a 1:1:2 mix of $e^+e^-$, $e^+e^-$, and $e^+e^-$ giving $e^+e^-$ and $e^+e^-$ fits of 0.82 and 0.51 to the AMS and Fermi data, and annihilation cross sections in the range $(6-23) imes 10^{-24}$ cm$^3$/s. Pulsars, modelled as a population with a birth rate of one per century and an $E^{-1.5}$-$E^{-1.6}$ injected spectrum cut off near 600 GeV, fit the same data at $e^+e^-$ and $e^+e^-$ if roughly 10-20 percent of their total energy goes into pairs, or only 3-4 percent if the nearby Geminga and B0656+14 pulsars dominate the local flux.","pith_inferences":["The paper's own Fig. 2 caption states that under the plain power-law background all the intermediate-state models are excluded, and the broken power law is introduced afterwards; a reader should treat the surviving fits as conditional on a background shape that was chosen partly to make them work, not as an independent detection.","Because the pulsar fit needs only a few percent of spin-down energy in pairs while the dark matter fit needs a thousand-fold cross-section boost, the two hypotheses are not equally natural, even though the paper presents them as equally viable at current sensitivity.","The Geminga and B0656+14 scenario makes a sharp prediction: the positron flux should be anisotropic at high energy, coming preferentially from the direction of those two sources, which is a test the paper does not attempt.","The same propagation uncertainty that rescues the dark matter fits also weakens any inference drawn from the total lepton spectrum, so a joint propagation-plus-source fit, rather than a scan over fixed $L$, would be the natural next step."],"forward_implications":["Any dark matter explanation of the positron excess must now proceed through intermediate states rather than direct leptonic annihilation, which sharply narrows the particle physics model space.","The required annihilation cross sections of roughly $10^{-24}$ to $10^{-23}$ cm$^3$/s remain two to three orders of magnitude above a thermal relic cross section, so some enhancement mechanism (Sommerfeld, non-thermal production, or clumping) is still mandatory.","Pulsars can supply the positrons with a few percent of their spin-down energy in pairs, which is within the range expected from polar-gap and outer-gap acceleration models.","A future AMS measurement of the $^{10}$Be/$^9$Be ratio at higher energy and precision would fix the diffusion zone thickness and thereby decide whether the surviving dark matter models live or die.","If the diffusion zone turns out to be narrow ($L \\lesssim 4$ kpc), the dark matter route closes and pulsars become the default explanation."],"supporting_citations":[{"why":"Supplies the AMS positron fraction measurement from 0.5 to 350 GeV that the entire re-analysis is fitted against.","marker":"[1]"},{"why":"GALPROP v54, the propagation code used to compute electron and positron spectra and energy losses.","marker":"[60]"},{"why":"Introduces exciting dark matter, the intermediate-state annihilation scenario that survives the new data.","marker":"[58]"},{"why":"Provides the analytic decay spectra used to compute the positron yields from the intermediate states.","marker":"[59]"},{"why":"Establishes the pulsar interpretation, including the Geminga and B0656+14 contribution, that the paper re-evaluates.","marker":"[22]"},{"why":"Supplies the secondary-to-primary ratio constraints used to fix the diffusion parameters and the allowed range of $L$.","marker":"[63]"},{"why":"Provides the spatial distribution of the pulsar birth rate adopted for the all-Milky-Way pulsar model.","marker":"[75]"},{"why":"A contemporaneous analysis of the same AMS data, noted as addressing similar issues.","marker":"[76]"}],"fun_headline_variants":["AMS positron shape disfavors direct leptonic dark matter annihilation","Only 1–3 TeV dark matter via short-lived intermediates still fits AMS","Nearby pulsars remain viable source of rising cosmic ray positrons","Core summary repeats e+e- labels, so surviving DM channels are unclear"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The surviving dark matter fits depend on a thick diffusion zone ($L = 8$ kpc) and on a primary electron spectrum that was made to harden above about 100 GeV; with a 2 kpc zone the same models give $e^+e^-$ per degree of freedom between 9.9 and 33, and the paper itself notes in the Fig. 2 caption that with the plain power-law background all of these models are excluded.","fun_headline_variants_meta":{"raw":{"variants":["AMS positron shape disfavors direct leptonic dark matter annihilation","Only 1–3 TeV dark matter via short-lived intermediates still fits AMS","Nearby pulsars remain viable source of rising cosmic ray positrons","Core summary repeats e+e- labels, so surviving DM channels are unclear"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2814,"prompt_tokens":1120,"completion_tokens":1694,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":1625}},"tokens_in":1120,"tokens_out":1694,"duration_ms":26032,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":1625,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T07:47:02.234355+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"An AMS measurement of the $^{10}$Be/$^9$Be ratio that pins the diffusion zone half-width below about 4 kpc would remove the propagation regime in which the surviving dark matter models fit, leaving pulsars as the only viable source among those considered here.","supporting_citations":[{"cited_title":"Physical Review Letters , keywords =","cited_arxiv_id":null,"evidence_quote":"Supplies the AMS positron fraction measurement from 0.5 to 350 GeV that the entire re-analysis is fitted against."},{"cited_title":"GALPROP Version 54: Explanatory Supplement","cited_arxiv_id":null,"evidence_quote":"GALPROP v54, the propagation code used to compute electron and positron spectra and energy losses."},{"cited_title":"Exciting Dark Matter and the INTEGRAL/SPI 511 keV signal","cited_arxiv_id":"astro-ph/0702587","evidence_quote":"Introduces exciting dark matter, the intermediate-state annihilation scenario that survives the new data."},{"cited_title":"High Energy Positrons and the WMAP Haze from Exciting Dark Matter","cited_arxiv_id":"0802.2922","evidence_quote":"Provides the analytic decay spectra used to compute the positron yields from the intermediate states."},{"cited_title":"Pulsars as the Sources of High Energy Cosmic Ray Positrons","cited_arxiv_id":"0810.1527","evidence_quote":"Establishes the pulsar interpretation, including the Geminga and B0656+14 contribution, that the paper re-evaluates."},{"cited_title":"Astrophysical Uncertainties in the Cosmic Ray Electron and Positron Spectrum From Annihilating Dark Matter","cited_arxiv_id":"0904.2398","evidence_quote":"Supplies the secondary-to-primary ratio constraints used to fix the diffusion parameters and the allowed range of $L$."},{"cited_title":"Birth and Evolution of Isolated Radio Pulsars","cited_arxiv_id":"astro-ph/0512585","evidence_quote":"Provides the spatial distribution of the pulsar birth rate adopted for the all-Milky-Way pulsar model."},{"cited_title":"Implications of the AMS-02 positron fraction in cosmic rays","cited_arxiv_id":"1304.1482","evidence_quote":"A contemporaneous analysis of the same AMS data, noted as addressing similar issues."}],"review_version":1}