{"id":"0aaf0ffc-a181-4692-b6c5-8b3b2044ef1b","arxiv_id":"2608.08144","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"For memory-burdened primordial black holes, a log-normal mass function can produce neutrino abundance limits several orders of magnitude stronger than a monochromatic population with the same median mass, with current IceCube data leading at low memory-burden strength and future radio detectors…","lead":"Memory-burdened primordial black holes could survive to today and emit neutrinos at energies from TeV to EeV. This paper computes how strongly current and future neutrino telescopes can constrain them, and shows that broad mass distributions are much easier to probe than single-mass populations.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The R90 enhancement may be a normalization effect: Eq. (47) compares log-normal vs monochromatic at equal characteristic mass, but Eq. (28) normalizes to present-day DM density, so the low-mass tail's steep per-PBH emission inflates the ratio; test k=0 or a matched initial-abundance comparison.","rationale":"The reader's weakest_assumption is the memory-burden ansatz of Eqs. (8)-(13). That is an external physics assumption, and the authors explicitly caveat it in Sec. II A, so it does not constitute an internal flaw. The present analysis focuses instead on the paper's own central quantitative claim: the orders-of-magnitude enhancement R90 for log-normal versus monochromatic populations. This claim is well-defined but depends on two choices: the comparison point (equal characteristic mass) and the normalization of fPBH (present-day DM density of survivors via Eq. (28)). Because the log-normal low-mass tail contains much lighter, hotter PBHs, its per-PBH emission rate is enhanced by a power of M^{-(2+2k)}; at the same present-day fPBH, the log-normal population therefore produces far more flux. The R90 numbers are a consequence of this steep mass scaling and the wide dynamic range of surviving masses, especially at larger k. My concrete test-evaluating R90 at k=0-would show whether the enhancement is a generic feature of extended mass functions or specifically amplified by memory burden. A matched-initial-abundance comparison would further separate the normalization effect from a genuine spectral-shape effect. This concern is not a rejection: the paper's frequentist limits, the IceCube vs. future-detector ordering, and the honest treatment of prior sensitivity are all solid. But the headline comparison metric needs an additional baseline before the 'several orders of magnitude' conclusion can be regarded as a robust physical statement rather than a convention-dependent one. The verdict remains CONDITIONAL, in agreement with the reader's overall assessment but for a different, more internal reason.","tokens_in":20303,"tokens_out":12255,"duration_ms":122287,"concrete_test":"Recompute the 90% CL abundance-limit ratio R90 of Eq. (47) for the standard Hawking case k=0, using the same present-day-mass normalization of Eq. (28) and the same log-normal widths sigma=0.5 and 1. If R90 is already extremely small (e.g., below 10^{-3}) over the corresponding mass range, the order-of-magnitude enhancement is generic to extended mass functions and not a memory-burden effect. Additionally, recompute R90 with both populations normalized to the same total initial mass density in the surviving window S_k instead of the same present-day DM density; if the matched-initial-abundance ratio is order unity while the present-day-normalized ratio is 10^{-6}, the headline claim is dominated by the normalization choice in Eq. (28).","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper is internally consistent and clearly caveated, so this is not a claim of error. The load-bearing issue is the definition of the headline quantity R90 in Eq. (47). The comparison is made 'at equal characteristic mass' (M0 = Mc), but the normalization in Eq. (28) ties the present-day weight W(M_i,k) to the present-day DM density of surviving PBHs. Because the log-normal distribution contains a low-mass tail of much lighter PBHs, those PBHs are far hotter and emit with a rate scaling roughly as M^{-(2+2k)} in the burdened phase. For the same fPBH, the log-normal population therefore produces dramatically more flux than a monochromatic population at the median mass. The resulting R90 values (10^{-6} for sigma=0.5, 10^{-9} for sigma=1) thus measure the dynamic range of surviving masses in the tail relative to the median, amplified by the steep mass scaling, rather than a robust statement about extended mass functions per se. The mechanism is exactly as the authors describe in Sec. IV B, but the headline numbers are a property of the chosen comparison baseline and the present-day-mass normalization. In particular, the same enhancement is expected to appear in the standard k=0 Hawking limit for any extended mass function with a tail below the median, which would mean the qualitative conclusion is not specific to memory-burdened PBHs. The paper's own prior-sensitivity check in Appendix C shows a related fragility: the strong Bayesian discrimination claim (ln B ~ 6.6) drops to ln B ~ 1.4 under a wider mass prior, so the 'orders of magnitude' language is not robust to reasonable alternative choices of how the comparison is set up.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript studies high- and ultra-high-energy neutrino signals from memory-burdened primordial black holes (PBHs), extending previous work by including IceCube MESE data and comparing monochromatic and log-normal initial mass functions. Using a two-stage Hawking-evaporation prescription with a memory-burden suppression factor S(M)^-k, the authors derive 90% CL upper limits on the present-day PBH abundance from current IceCube HESE, MESE, and EHE data, and projected limits for IceCube-Gen2 radio and GRAND200k. They define a ratio R90 comparing log-normal and monochromatic limits at equal characteristic mass and find that the log-normal limits can be stronger by several orders of magnitude. They then run Bayesian 30-event forecasts for combined future detectors, computing Bayes factors between the two mass-function hypotheses and posterior parameter reconstructions, with prior-sensitivity tests in Appendix C.","tokens_in":20697,"tokens_out":17287,"duration_ms":190994,"significance":"If the adopted memory-burden prescription is correct, the paper provides new leading abundance constraints (especially through the first inclusion of MESE data) and makes a useful cautionary point: extended mass functions, not just monochromatic ones, should be used when interpreting PBH limits. The analysis is transparent: flux formulas, likelihood constructions, exposure treatments, and priors are stated explicitly; the Bayesian forecasts are clearly signal-only, and the prior dependence of the model-comparison evidence is tested and reported. The derivations are checkable, and the paper does not overclaim beyond its stated assumptions. The main limitations (instantaneous transition at q = 1/2, signal-only forecasts, no energy-resolution smearing) are acknowledged in the text.","major_comments":[{"comment":"The headline R90 values are strongly shaped by the normalization in Eq. (28) and by the steep mass dependence of the emission rate in the burdened phase. Because the limits are normalized to the present-day DM density of surviving PBHs, a log-normal population with the same median mass and the same fPBH automatically contains more low-mass, high-temperature PBHs than a monochromatic population, and the per-PBH emission in the burdened phase scales roughly as M^{-(2+2k)}. The paper states this mechanism correctly, but the headline factors (R90 ~ 10^-6 for sigma = 0.5 and 10^-9 for sigma = 1) are therefore a property of the chosen comparison baseline rather than a model-independent measure of the constraining power of extended mass functions. To make this central quantitative claim robust, please add a k = 0 control case or a matched-initial-abundance comparison (e.g., equal total initial PBH mass density), and explicitly discuss in Sec. IV B how much of the enhancement is specific to memory burden rather than a generic consequence of the log-normal tail.","section":"Sec. IV B, Eq. (47)"}],"minor_comments":[{"comment":"The text contains the typos \"tentalizing\" (abstract) and \"instanateous\" (Sec. II A); these should read \"tantalizing\" and \"instantaneous\".","section":"Abstract and Sec. II A"},{"comment":"In the first paragraph, \"The strongest upper limit is obtained for k = 1 near a characteristic mass of 3e7 g and are approximately\" has a subject-verb mismatch; it should read \"is approximately\".","section":"Sec. IV A"},{"comment":"The sentence \"these results significantly depends on the choice of prior\" should read \"depend\".","section":"Sec. IV C"},{"comment":"Please define the units of n_N explicitly (cm^-3) and state that the effective volumes of Ref. [53] are in cm^3, so that A_eff in Eq. (37) is obtained in cm^2; this will improve reproducibility.","section":"Sec. III B, Eq. (37)"},{"comment":"The quoted R90 minima (10^-6 for sigma = 0.5 and 10^-9 for sigma = 1) correspond to the upper ends of the displayed mass ranges; the figure caption or text should state the exact characteristic masses at which these minima occur.","section":"Fig. 3 and Sec. IV B"},{"comment":"The reduction of the combined median lnB from 6.60 to 1.42 under the wider mass prior is an important result; consider showing the distribution of lnB for the widened prior alongside the baseline in a figure or table to make the prior sensitivity more transparent.","section":"Appendix C"}],"recommendation":"major_revision","confidential_remarks":"The paper is within scope for this journal and is technically sound. The main issue I would like the authors to address before publication is the interpretation of the R90 enhancement: as written, the headline orders-of-magnitude ratio partly reflects the present-day normalization in Eq. (28) and the steep emission scaling in the memory-burdened phase, not only the difference between mass-function models. The authors are transparent about this, but a k = 0 control case or a matched-initial-abundance comparison would make the central claim much more robust. I do not see grounds for rejection; the prior sensitivity of the Bayesian evidence is already disclosed and is not an error."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read the paper with the memory-burden PBH neutrino limits. The genuinely new pieces are the first IceCube MESE constraints on these PBHs and the first systematic monochromatic vs log-normal comparison, including Bayesian model-discrimination forecasts. On the technical side, it is well executed. The flux formulas, one-sided likelihood, and Feldman-Cousins forecasts are explicit; the background-agnostic treatment of the IceCube data is sensible; and Appendix C does a real prior-sensitivity check. The authors also disclose the idealizations in the Bayesian part—signal-only simulations, no energy-resolution smearing, detector-motivated priors—rather than burying them. That is good practice.\n\nThe soft spot is the headline ratio R90. In Eq. (47), they compare log-normal and monochromatic populations at equal characteristic mass, with both normalized to the same present-day f_PBH. Because the log-normal has a low-mass tail, and emission scales steeply with mass, the tail dominates the flux, so the log-normal limit is stronger by orders of magnitude. The authors state this mechanism clearly. But I think they underplay that the same effect would appear in the standard k=0 Hawking case for any extended mass function with a tail below the median. The R90 values are therefore mostly a statement about the dynamic range of the mass function, not about memory burden. A quick k=0 test would settle it, and if it does, the 'orders of magnitude' language in the abstract and conclusions should be reframed accordingly. The prior sensitivity in Appendix C matters here too: the log-normal Bayes factor drops from ~6.6 to ~1.4 under a wider mass prior, so the strong-discrimination claim is conditional on a fairly local prior choice. To their credit, they report this, but the abstract reads as if the discrimination story is stronger than it is.\n\nMinor: no analysis scripts or data products appear to be released, which makes the forecasts harder to reproduce. The memory-burden ansatz itself is inherited from the literature; the authors flag the smooth-transition caveat, so I don't hold that against the paper.\n\nI would send this to a serious referee. It is a solid, competent piece of phenomenology that extends the previous monochromatic analysis. The R90 framing needs work, and reproducibility would help, but the constraints and forecasts are worth having on record.","headline":"A careful, honest phenomenology paper whose headline R90 mass-function comparison says more about the low-mass tail of extended distributions than about memory burden—still worth a serious referee.","tokens_in":21191,"tokens_out":3214,"would_cite":true,"duration_ms":32958,"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":"A log-normal spread of black-hole masses can strengthen neutrino-telescope limits on memory-burdened primordial black holes by up to $10^{-9}$ at fixed median mass.","keywords":["primordial black holes","memory burden","Hawking radiation","high-energy neutrinos","IceCube","IceCube-Gen2 radio","GRAND200k","log-normal mass function"],"falsifier":"Take the benchmark log-normal point ($M_c=3.16\\times10^{5}$ g, $\\sigma=1$, $k=2$, $f_{\\mathrm{PBH}}=8.22\\times10^{-8}$) that is predicted to yield 30 events in ten years of combined IceCube-Gen2 radio and GRAND200k; if the combined observed count is below the background-free 90% upper limit of 2.44 events, that benchmark and the suppression law underlying it are excluded.","tokens_in":20120,"feed_emoji":"🕳️","tokens_out":8121,"duration_ms":77707,"temperature":0.7,"pith_summary":"Primordial black holes whose evaporation is slowed by the memory-burden effect can survive to the present at masses below the usual Hawking threshold and emit TeV-to-EeV neutrinos. This paper asks how strongly current and planned neutrino telescopes can constrain such black holes, and whether the answer depends on whether the black-hole masses are all equal or spread out. It finds that a log-normal spread can make the neutrino abundance limits several orders of magnitude stronger than the monochromatic approximation at the same median mass, sometimes by a factor of $10^{-6}$ to $10^{-9}$. The reason is the low-mass tail of the distribution: lighter black holes are hotter, so the tail dominates the high-energy neutrino flux. If true, this means that monochromatic studies of memory-burdened black holes can understate the discovery reach by many orders of magnitude.","feed_headline":"Black-hole mass spread sharpens neutrino limits a billionfold","feed_subtitle":"A wider black-hole mass distribution means neutrino telescopes can probe far smaller black-hole abundances.","key_machinery":"The machinery is a two-stage evaporation prescription: a black hole evaporates by standard Hawking physics until its mass falls to $M_q=qM_i$ with $q=1/2$, then enters a memory-burdened phase in which both neutrino emission and mass loss are suppressed by the entropy factor $S[M]^{-k}$, where $S(M)=4\\pi G M^2$ and $k$ measures the memory-burden strength. This factor lets lighter black holes survive to the present and gives the burdened population its high-energy neutrino spectrum. Superimposed on it is the initial mass function: a monochromatic delta function versus a log-normal distribution. The log-normal's low-mass tail, containing lighter surviving black holes with higher Hawking temperatures, is what produces the stronger limits and the broader, higher-energy flux.","core_discovery":"The paper's central claim is that, within the memory-burden evaporation model, the initial black-hole mass function is not a detail: it changes the neutrino-based constraints on the present-day black-hole abundance by orders of magnitude. Treating the population as log-normal with width $\\sigma=0.5$ or $1$ strengthens the strongest 90% confidence abundance limit relative to a monochromatic population at equal characteristic mass, with the ratio $R_{90}$ dropping to roughly $10^{-6}$ for $\\sigma=0.5$ and $10^{-9}$ for $\\sigma=1$ over the masses considered. Current IceCube data give the leading limits for weak memory burden ($k=1$), reaching $f_{\\mathrm{PBH}}\\simeq10^{-9}$ for the monochromatic case and $9\\times10^{-11}$ for the log-normal case near $3\\times10^{7}$ g, while projected IceCube-Gen2 radio and GRAND200k become the strongest probes for stronger burden ($k=2,4$), with the best projected limit $f_{\\mathrm{PBH}}\\simeq3\\times10^{-12}$ for a log-normal population near $10^{4}$ g. A simulated 30-event signal would distinguish a broad log-normal distribution from a monochromatic one with very strong evidence for $k_{\\mathrm{true}}=2$ under the baseline prior, but not reliably for $k_{\\mathrm{true}}=4$.","pith_inferences":["Beyond the paper, the same low-mass-tail mechanism should also enhance gamma-ray and cosmic-ray constraints on memory-burdened PBHs, so extended-mass limits in other channels may be similarly stronger than monochromatic estimates.","Beyond the paper, one testable extension is to fold the flux predictions into a combined multi-detector likelihood: because each detector probes a different energy slice, a single log-normal population with fixed $(M_c,\\sigma,k)$ predicts a specific pattern of detections and non-detections across IceCube, IceCube-Gen2 radio, and GRAND200k that the paper does not explicitly combine.","Beyond the paper, the reported sensitivity of the Bayesian discrimination to the prior width, especially the reduction of the $k_{\\mathrm{true}}=2$ log-normal Bayes factor from 6.60 to 1.42 under a wider mass prior, suggests that a data-driven prior anchored to a specific formation model would be needed before a real detection could be claimed as evidence for a broad mass function.","Beyond the paper, if the memory-burden transition is smooth rather than instantaneous, the effective suppression starts earlier and the surviving mass window changes; applying the same analysis with the smooth-transition models cited in the paper would likely shift the strongest mass ranges by a non-negligible amount."],"forward_implications":["Current IceCube HESE, MESE, and EHE data already set the leading PBH abundance limits for $k=1$, and projected radio detectors do not improve on them in this regime.","For $k=2$ and $k=4$, the projected IceCube-Gen2 radio and GRAND200k exposures give substantially stronger limits than current IceCube data, because surviving black holes are lighter and emit at higher energies.","Evaluating a monochromatic constraint at the median mass of a log-normal population underestimates the true limit by up to several orders of magnitude, so monochromatic-only studies should be read with caution.","A future 30-event signal from combined IceCube-Gen2 radio and GRAND200k would likely identify a broad log-normal distribution as the source for $k_{\\mathrm{true}}=2$, but would not reliably distinguish mass functions for $k_{\\mathrm{true}}=4$.","Parameter reconstruction works well for a monochromatic population at $k=2$ ($M_0$ and $k$ to roughly ten percent) but degrades for log-normal populations, especially at higher $k$, where $M_c$, $k$, and $\\sigma$ trade off against one another."],"supporting_citations":[{"why":"Introduces memory-burden stabilization of black holes, the physical effect that lets sub-Hawking-mass PBHs survive.","marker":"[17]"},{"why":"Extends the memory-burden framework to black holes and solitons, providing the theoretical basis for the suppressed evaporation stage.","marker":"[18]"},{"why":"Gives the previous IceCube EHE/HESE plus Gen2/GRAND constraints and the background-agnostic likelihood method reused here.","marker":"[27]"},{"why":"IceCube EHE event-count data serve as a current high-energy constraint.","marker":"[25]"},{"why":"IceCube 7.5-year HESE flux data serve as a current constraint.","marker":"[26]"},{"why":"IceCube MESE data are incorporated for the first time as an extra current constraint.","marker":"[28]"},{"why":"IceCube-Gen2 radio exposure defines the projected sensitivity forecasts.","marker":"[23]"},{"why":"GRAND200k exposure defines the projected sensitivity forecasts.","marker":"[24]"},{"why":"Argues monochromatic constraints cannot be mapped to extended mass functions, motivating the log-normal comparison.","marker":"[33]"},{"why":"Provides the extended-mass-function constraint formalism this paper uses to define the log-normal limits.","marker":"[34]"}],"fun_headline_variants":["Log-normal black-hole masses tighten neutrino limits a billionfold","Mass-shape twist sharpens neutrino probes of black holes","PBH mass distribution keys neutrino telescope sensitivity","Wider black-hole masses sharpen neutrino abundance limits","Neutrino bounds on black holes hinge on mass spread"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the two-stage memory-burden prescription: a black hole evaporates normally until it loses half its mass, then emission is suddenly suppressed by the entropy factor $S[M]^{-k}$ with $k$ fixed; if the real transition is gradual or the suppression follows a different law, every limit and forecast shifts.","fun_headline_variants_meta":{"raw":{"variants":["Log-normal black-hole masses tighten neutrino limits a billionfold","Mass-shape twist sharpens neutrino probes of black holes","PBH mass distribution keys neutrino telescope sensitivity","Wider black-hole masses sharpen neutrino abundance limits","Neutrino bounds on black holes hinge on mass spread"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00055,"raw_usage":{"total_tokens":2708,"prompt_tokens":1108,"completion_tokens":1600,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":724,"completion_tokens_details":{"reasoning_tokens":1524}},"tokens_in":724,"tokens_out":1600,"duration_ms":14321,"temperature":1.0,"reasoning_tokens":1524,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:20:47.985868+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the benchmark log-normal point ($M_c=3.16\\times10^{5}$ g, $\\sigma=1$, $k=2$, $f_{\\mathrm{PBH}}=8.22\\times10^{-8}$) that is predicted to yield 30 events in ten years of combined IceCube-Gen2 radio and GRAND200k; if the combined observed count is below the background-free 90% upper limit of 2.44 events, that benchmark and the suppression law underlying it are excluded.","supporting_citations":[{"cited_title":"Probing Memory-Burdened Primordial Black Holes with High-Energy Neutrinos","cited_arxiv_id":"2608.08144","evidence_quote":"GRAND200k exposure defines the projected sensitivity forecasts."}],"review_version":1}