{"id":"ce539590-60b0-4ea5-b984-79c089b28f25","arxiv_id":"2505.10451","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"N-body simulations and rotation-curve data constrain the dark matter self-interaction cross-section to σ/m ≤ 2.7 cm²/g at 95% C.L. for Milky Way-scale halos.","lead":"This paper simulates dark matter halos with self-interactions and finds that the most probable dark matter speed rises with the self-interaction cross-section. Comparing this rise with measured rotation speeds in Milky Way-like galaxies yields an upper limit of 2.7 cm² per gram, adding a new observable to the dark matter self-interaction toolkit.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The σ/m≤2.7 bound is calibrated on NFW halos with c≈100, far denser than MW-like halos; this makes the limit stronger, not conservative.","rationale":"The reader identified the v0(R⊙) ≈ v⊙ mapping as the weakest assumption. My concern is closely related but more specific and, in my reading, more load-bearing: the mapping is applied to halos whose internal structure is far from the Milky Way's. The Table 1 initial conditions imply concentration c ≈ 100 for every halo. Such halos are much denser in the inner region than a real MW-mass halo, so at fixed σ/m the self-interaction rate and the resulting change in v0 are exaggerated. The text's claim that ignoring baryons makes the limits conservative only addresses one direction of bias; it does not account for the density normalization of the simulated halos. A conservative bound should be robust to moving from c ≈ 100 to the cosmologically expected c ≈ 10–15, and the paper does not provide any such check. I also note that the central 2.7 cm²/g value is not easily reproduced from the quoted P2 coefficients and the stated 233 ± 6 km/s at 95% C.L.; a full likelihood or confidence-interval construction is not shown. That reproducibility issue is real but secondary: even if the arithmetic were clarified, the bound would still inherit the calibration-sample problem. I therefore recommend CONDITIONAL acceptance: the authors should demonstrate that the simulated halos reproduce the MW rotation curve and that the v0–σ/m response is stable under a realistic concentration, or re-derive the limit with such halos.","tokens_in":16572,"tokens_out":20869,"duration_ms":221503,"concrete_test":"Rerun the Milky-Way-scale analysis (Section 5.1) for M = 10^12 M⊙ with a cosmologically motivated NFW concentration c ≈ 10, corresponding to r_s ≈ 20 kpc rather than the r_so ≈ 2 kpc in Table 1, keeping the SIDM implementation, halo mass definition, fitting procedure, and observed 233 ± 6 km/s band unchanged. If the recomputed 95% C.L. upper limit on σ/m shifts above 5 cm²/g (or changes by more than a factor of two), the quoted 2.7 cm²/g limit is an artifact of the extreme c ≈ 100 initial conditions and is not conservative.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The headline bound rests on the empirical relation v0(σ/m, M) extracted from the simulations in Table 1. Every halo is set up with r_cut/r_so ≈ 100, i.e. an NFW concentration c ≈ 100. This is far outside the standard CDM concentration–mass relation for Milky-Way-mass halos, where c ≈ 10–15. Since the SIDM heating rate inside the halo scales with the local DM density, these overdense halos will produce a larger increase of the MB most probable speed v0 with σ/m than a realistic MW halo would. The paper's only conservatism argument (Section 5.1: baryons 'aid thermalization') does not address this: high concentration makes the inferred limit stronger, not weaker. For M = 10^12 M⊙ with r_so ≈ 2 kpc, the DM-only NFW circular velocity at R = 8.2 kpc is of order 300–350 km/s, far above the observed 233 ± 6 km/s, while the fitted v0 used in Figure 6 is only about 230 km/s. Thus the simulated halos do not reproduce the MW rotation curve, and the quantity v0 identified in the analysis is not the observed circular velocity in these halos. The claimed 2.7 cm²/g bound is therefore not established as a conservative constraint on real MW-size halos; it may be artificially tight because of the extreme initial conditions.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper uses GADGET N-body simulations of isolated NFW halos with velocity-independent elastic dark-matter self-interactions to study how the most probable speed v0 of a Maxwell-Boltzmann fit to the DM speed distribution depends on halo mass and on the specific cross section σ/m. The authors fit two empirical forms, a polynomial (P1) and a power law (P2), to simulated v0 values using MCMC, then compare the resulting v0(σ/m) at a benchmark Milky-Way mass with the observed solar circular velocity v⊙ ≈ 233 ± 6 km/s. From this comparison they derive σ/m ≤ 2.7 cm²/g (P2) and ≤ 3.5 cm²/g (P1) at 95% C.L., with weaker bounds for the LSB galaxy F563-1 and the cluster A611. The central claim is a conservative upper limit on velocity-independent SIDM from rotation-curve observations.","tokens_in":16853,"tokens_out":13147,"duration_ms":132218,"significance":"Should the bound hold, it would add a competitive, velocity-distribution-based constraint comparable to Bullet Cluster limits, with the advantage of applying to Milky-Way-mass halos. The paper contains a systematic simulation campaign across nine halo masses and eight cross sections with four realizations each, and it provides a useful quantitative comparison of Maxwell-Boltzmann, Gaussian, and Tsallis fits to simulated velocity distributions. However, the headline bound is not yet established: the benchmark mass is internally inconsistent, the simulated halos are extremely concentrated compared with Milky-Way-like systems, and the assumed v0 ≈ v⊙ mapping is not satisfied by the simulations themselves. These are load-bearing issues rather than presentation defects.","major_comments":[{"comment":"The headline bound cannot be reproduced from the published coefficients at the stated benchmark mass. For the P2 fit in Table 2 with M_halo = 8×10^11 M⊙ (i.e., y = 80 in Eq. 4.2), v0(σ/m=0) = 221.9 km/s and v0(σ/m=5 cm²/g) = 241.0 km/s, so the 95% upper edge of 245 km/s is crossed only at σ/m ≈ 7 cm²/g, not at 2.7 cm²/g. The quoted 2.7 cm²/g corresponds instead to y = 100 (M_halo = 10^12 M⊙), for which v0(0) = 231 km/s and the crossing occurs at σ/m ≈ 2.8 cm²/g. Similarly, the P1 fit gives a crossing near 5.2 cm²/g for y = 80 but near 3.8 cm²/g for y = 100, while the text quotes 3.5 cm²/g. The stated benchmark mass of 8×10^11 M⊙ is therefore inconsistent with the plotted curves and the quoted bounds. The authors must state the exact benchmark mass and radius used in Figure 6, tabulate the v_sim_o values extracted from the simulations, and show the intersection calculation explicitly.","section":"§5.1, Table 2, Fig. 6"},{"comment":"All simulated halos are initialized with r_cut/r_so ≈ 100, i.e., an NFW concentration c ≈ 100. For the 10^12 M⊙ halo, r_so = 1.99 kpc and r_cut = 199 kpc. The standard CDM concentration–mass relation for this mass gives c ≈ 10–15, so the simulated halos are far denser than Milky-Way-like halos. Concretely, the c = 100, 10^12 M⊙ NFW halo has a DM-only circular velocity of about 346 km/s at r = 8.2 kpc, well above the observed 233 ± 6 km/s; the same halo gives about 338 km/s at r = 10 kpc. Because the SIDM heat-transfer rate scales with the local DM density, these overdense initial conditions inflate the response of v0 to σ/m, making the derived limit artificially strong rather than conservative. The argument in §5.1 that baryons 'aid thermalization' does not address this concentration bias. The authors need to recalibrate with realistic concentrations, or demonstrate quantitatively that the v0(σ/m) relation is insensitive to c.","section":"Table 1 and §5.1"},{"comment":"The load-bearing mapping v0(R⊙) ≈ v⊙ is not satisfied by the authors' own simulations. In the c = 100, 10^12 M⊙ halo, the circular velocity at r = 10 kpc is ≈ 338 km/s, while the fitted Maxwell-Boltzmann v0 in Figure 5 is ≈ 230 km/s. Thus v0 ≠ v_circ within the simulated halos, and comparing a simulation-derived v0 directly to the observed stellar circular velocity is not self-consistent unless the ratio v0/v_circ is modeled. In addition, the analysis assumes the MB form even though Figure 3 shows that the Gaussian gives a better fit, and it uses the same v⊙ measurement both to set the MCMC priors and to read off the final bound. The authors should marginalize over the velocity-distribution shape and over the v0/v_circ mapping, or provide a clear physical justification for the adopted identification.","section":"§5.1, Eq. (3.1), Fig. 3"}],"minor_comments":[{"comment":"The columns labeled 'Prior Derived' are ambiguous, and several derived values (e.g., b2 = 10^-4 for the Milky-Way and 5×10^-3 for the cluster) are quoted without uncertainties; the log-scaling note in §5.3 is not clearly reflected in the table.","section":"Table 2"},{"comment":"The caption contains a typo: '44πv2f(v)' should read '4πv²f(v)'.","section":"Figure 1 caption"},{"comment":"References [9] and [10] appear internally mislabeled: [9] is listed as Flores & Primack with an incomplete journal citation, while [10] is attributed to 'Planck collaboration, A. Burkert' with a 1995 journal entry; these entries should be corrected.","section":"References"},{"comment":"The footnote describing the Maxwell-Boltzmann fit to simulated data does not state the number of velocity bins, the weighting scheme, or how the errors used for χ²_r are obtained; these details are needed to judge the fit-quality comparison in Figures 2 and 3.","section":"Section 3 footnote"},{"comment":"The text says 'the value of v0 at a galactocentric distance of 8.2 kpc is measured to be (233±6) km/s', but 233±6 km/s is the measured circular velocity v⊙, not a measured v0; the notation should distinguish the observed quantity from the model parameter throughout.","section":"§5.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is largely an extension of the authors' earlier JCAP 2022 core-size analysis, and the new velocity-distribution-based constraint is the main addition. The central issues identified above are technical and appear fixable by a reanalysis with realistic concentrations and a transparent benchmark choice, but in the present form the headline bound is not established."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThis paper tries to turn the velocity distribution of dark matter in halos into an SIDM constraint. The authors run a substantial set of isolated N-body simulations (nine masses, eight cross-sections, four realizations), fit the local speed distribution to a Maxwell-Boltzmann form, and parameterize the most probable speed v0 as a function of σ/m and halo mass. They then compare the fitted v0 to observed rotation velocities and claim a conservative bound σ/m ≤ 2.7 cm²/g at 95% C.L. from Milky-Way-size halos.\n\nWhat's genuinely new is the parameterization v0(σ/m, M). The closest prior is their own core-size paper [37], and I don't know of another group using this observable this way. The simulation work is standard and reasonably careful, though I haven't audited the convergence tests.\n\nBut the central result does not hold up to inspection. First, the paper doesn't show how the 95% C.L. intersection is computed. More troubling, the fit coefficients in Table 2 don't reproduce the quoted 2.7 cm²/g when I plug in the benchmark mass of 8×10^11 M⊙; the power-law gives a crossing around 4 cm²/g, and the polynomial around 4.5. That suggests the quoted number comes from a different calculation or there's an error.\n\nSecond, and more serious, the simulated halos are initialized with r_cut/r_so ≈ 100, i.e., NFW concentrations around 100. For a 10^12 M⊙ halo with r_so ≈ 2 kpc, the CDM circular velocity at 8.2 kpc is ~350 km/s, far above the observed 233 km/s. These halos are not Milky-Way-like at the solar radius. Because the SIDM heating rate scales with density, the v0(σ/m) relation from these overdense halos will overestimate the effect for a given cross-section, making the bound tighter than it should be. The paper calls the bound conservative because baryons would aid thermalization, but the concentration issue points the other way. The two effects might partially cancel, but the paper does not demonstrate that.\n\nThe mapping v0(R_sun) ≈ v_sun is also a crude step. The observed v_sun includes baryons; in a DM-only simulation, v0 is the DM circular velocity, which is lower. That makes the comparison at least ambiguous.\n\nIn short: the idea is interesting and the simulations are a lot of work, but the headline bound is not reproducible and the halo setup undermines the claim of conservatism. The paper would benefit from a transparent derivation, a realistic halo concentration or a check that the result is insensitive to c, and a comparison to their earlier core-size bound.\n\nI would send this to peer review: the authors should be asked to fix these issues, and there is enough substance to justify referee time. But I would not cite the 2.7 number in its current form.","headline":"The headline σ/m bound is not reproducible from the paper's own coefficients, and the simulated halos are far more concentrated than real Milky-Way-size halos, so the claim of conservatism is doubtful.","tokens_in":17448,"tokens_out":12238,"would_cite":false,"duration_ms":104362,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Self-interacting dark matter with a cross-section above about 2.7 cm²/g would distort the local dark-matter velocity distribution more than Milky Way rotation-curve observations allow.","keywords":["dark matter self-interaction","velocity distribution function","Maxwell-Boltzmann distribution","rotation curves","Milky Way halo","low surface brightness galaxies","galaxy clusters","SIDM constraints"],"falsifier":"Measure the local dark-matter velocity distribution well enough to fix $v_0(R_\\odot)$ independently of the rotation curve, for example with a directional direct-detection experiment or reconstruction from tidal streams. If the most probable speed at 8.2 kpc falls outside the $233\\pm 6$ km/s band once the full velocity ellipsoid is modeled, the paper's central mapping fails; in that case, re-running the analysis with the measured distribution should shift or erase the $\\sigma/m \\le 2.7\\,\\mathrm{cm}^2/\\mathrm{g}$ bound.","tokens_in":16296,"feed_emoji":"🌌","tokens_out":10706,"duration_ms":70918,"temperature":0.7,"pith_summary":"This paper tries to establish that dark-matter self-interactions leave a measurable imprint in the velocity distribution of dark matter inside galactic halos, not only in their density. From N-body simulations of isolated halos spanning low-surface-brightness galaxies, Milky-Way-sized spirals, and galaxy clusters, the authors find that self-scattering raises the most probable dark-matter speed inside the thermalized core, and they encode this shift as a function of halo mass and cross-section. Comparing the predicted shift with rotation-curve observations at the Solar circle yields a conservative upper bound $\\sigma/m \\le 2.7\\,\\mathrm{cm}^2/\\mathrm{g}$ at 95% confidence for Milky-Way-like halos, with weaker bounds of about $9.8\\,\\mathrm{cm}^2/\\mathrm{g}$ for an LSB galaxy and about $10\\,\\mathrm{cm}^2/\\mathrm{g}$ for a cluster. A sympathetic reader would care because a single, relatively local observable, the speed of stars in a spiral galaxy, would then constrain a fundamental particle property of dark matter.","feed_headline":"Galaxy rotation curves cap dark-matter self-scattering at 2.7 cm²/g","feed_subtitle":"Self-scattering shifts dark-matter speeds; Milky Way rotation data cap σ/m at 2.7 cm²/g.","key_machinery":"The mechanism that carries the argument is the most probable speed $v_0$ of a truncated Maxwell-Boltzmann velocity distribution, $f(v)\\propto \\exp(-|v|^2/v_0^2)$ for $|v|\\le v_{\\rm esc}$. The paper treats $v_0$ as a fitted function of halo mass and $\\sigma/m$, calibrated to simulations with two empirical forms, P1 (polynomial) and P2 (power law). Self-scattering redistributes energy and pushes $v_0$ upward in the core; the observational bridge is the approximation $v_0(R_\\odot)\\simeq v_\\odot$, in which collisionless stars are treated as tracers of the dark-matter dynamics. Comparing the simulated rise in $v_0$ with the measured stellar rotation speed converts rotation-curve data into an upper bound on $\\sigma/m$.","core_discovery":"The paper's central claim is that the deformation of the dark-matter velocity distribution from self-interaction is captured by an increase in the most probable velocity $v_0$ of a truncated Maxwell-Boltzmann fit, and that this increase can be calibrated against rotation-curve data to bound the cross-section. After fitting $v_0(\\sigma/m, M_{\\rm halo})$ with both a polynomial and a power-law ansatz to SIDM N-body simulations, the authors compare the predicted $v_0$ at the Solar radius with the observed local circular velocity $v_\\odot = 233\\pm 6$ km/s at 8.2 kpc. Requiring consistency at 95% confidence gives $\\sigma/m \\le 2.7\\,\\mathrm{cm}^2/\\mathrm{g}$ for the power-law fit and $\\sigma/m \\le 3.5\\,\\mathrm{cm}^2/\\mathrm{g}$ for the polynomial fit. Repeating the same procedure for the LSB galaxy F563-1 and the cluster A611 gives $\\sigma/m \\le 9.8\\,\\mathrm{cm}^2/\\mathrm{g}$ and $\\sigma/m \\sim 10\\,\\mathrm{cm}^2/\\mathrm{g}$, respectively. The authors emphasize that the Milky-Way bound is conservative because baryonic effects, which would strengthen thermalization, are omitted from the simulations.","pith_inferences":["A direct kinematic measurement of the local dark-matter velocity distribution, from tidal streams or from the annual modulation signal in a directional detector, could test the mapping $v_0(R_\\odot)\\simeq v_\\odot$; a deviation larger than the quoted $6$ km/s band would shift the bound.","The authors report that the generalized Gaussian fits the simulated distributions slightly better than the Maxwell-Boltzmann form, so re-deriving the bound with a Gaussian or Tsallis shape would show how much of the constraint depends on the assumed functional form.","The isolated-halo simulations neglect the mergers and accretion that build halos in cosmological settings; if hierarchical assembly changes the inner velocity distribution, the calibrated $v_0(\\sigma/m)$ relation, and hence the bound, could move.","The method could be applied to the next round of stellar kinematic surveys: more galaxies with high-resolution inner rotation curves would turn this one-object bound into a population-level test of velocity-dependent self-interactions."],"forward_implications":["Velocity-independent, elastic dark-matter self-interactions with $\\sigma/m \\gtrsim 2.7\\,\\mathrm{cm}^2/\\mathrm{g}$ are excluded for Milky-Way-like halos at 95% confidence under the Maxwellian velocity assumption used here.","The same pipeline yields scale-dependent bounds: $\\sigma/m \\lesssim 9.8\\,\\mathrm{cm}^2/\\mathrm{g}$ for LSB galaxies with inner rotation curves like F563-1 and $\\sigma/m \\lesssim 10\\,\\mathrm{cm}^2/\\mathrm{g}$ for relaxed clusters like A611.","Because baryonic heating is omitted, the reported limits are one-sided and conservative; adding baryons would only strengthen the case for small $\\sigma/m$.","Within the thermalized core, self-interaction makes the Maxwell-Boltzmann form a progressively better description of the simulated velocity distribution as $\\sigma/m$ grows, supporting the use of quasi-thermal distributions for SIDM halos.","Direct-detection and indirect-detection analyses that assume a fixed $v_0\\simeq v_\\odot$ should treat $v_0$ as $\\sigma/m$-dependent when interpreting signals from self-interacting models."],"supporting_citations":[{"why":"Provides the scattering algorithm the simulations adopt for velocity-independent elastic self-interactions.","marker":"[36]"},{"why":"Establishes the simulation setup and stability criteria that this work extends to nine halo masses, and is the predecessor whose core-size constraints motivate the velocity-distribution analysis.","marker":"[37]"},{"why":"Defines the critical radius within which self-interactions thermalize the core, justifying the Solar neighborhood and the cluster mean-velocity relation used to set priors.","marker":"[54]"},{"why":"Supplies the standard-halo-model Maxwell-Boltzmann form and the approximation that collisionless stars trace the dark-matter velocity, $v_0(R_\\odot)\\simeq v_\\odot$.","marker":"[59]"},{"why":"Provides the Milky Way rotation-curve data at the Solar position used as the observational anchor for the main bound.","marker":"[81]"},{"why":"Gives the measured local circular velocity $233\\pm 6$ km/s at 8.2 kpc that fixes the prior and the 95% comparison band.","marker":"[86]"},{"why":"Supplies the F563-1 rotation-curve data, at $60.89\\pm 7.76$ km/s at 2.26 kpc, used for the LSB bound.","marker":"[95]"},{"why":"Provides the A611 cluster kinematics, 236 galaxies with measured velocity dispersion, used to fix the cluster-scale prior.","marker":"[106]"}],"fun_headline_variants":["Dark-matter speeds cap self-scattering at 2.7 cm²/g","Rotation curves set 2.7 cm²/g limit on dark-matter self-interaction","Velocity distribution bounds dark-matter self-scattering to 2.7 cm²/g","SIDM constrained: σ/m ≤ 2.7 cm²/g from galactic rotation","Dark-matter self-scattering limit from velocity profiles"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the identification of the dark matter's most probable speed at the Sun's position with the measured stellar circular velocity, $v_0(R_\\odot)\\simeq v_\\odot=233\\pm 6$ km/s; if the dark-matter velocity distribution is not the assumed bell-shaped form, or is direction-dependent, the reported bound no longer follows.","fun_headline_variants_meta":{"raw":{"variants":["Dark-matter speeds cap self-scattering at 2.7 cm²/g","Rotation curves set 2.7 cm²/g limit on dark-matter self-interaction","Velocity distribution bounds dark-matter self-scattering to 2.7 cm²/g","SIDM constrained: σ/m ≤ 2.7 cm²/g from galactic rotation","Dark-matter self-scattering limit from velocity profiles"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000179,"raw_usage":{"total_tokens":1293,"prompt_tokens":933,"completion_tokens":360,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":549,"completion_tokens_details":{"reasoning_tokens":255}},"tokens_in":549,"tokens_out":360,"duration_ms":3948,"temperature":1.0,"reasoning_tokens":255,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:10:04.995074+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the local dark-matter velocity distribution well enough to fix $v_0(R_\\odot)$ independently of the rotation curve, for example with a directional direct-detection experiment or reconstruction from tidal streams. If the most probable speed at 8.2 kpc falls outside the $233\\pm 6$ km/s band once the full velocity ellipsoid is modeled, the paper's central mapping fails; in that case, re-running the analysis with the measured distribution should shift or erase the $\\sigma/m \\le 2.7\\,\\mathrm{cm}^2/\\mathrm{g}$ bound.","supporting_citations":[],"review_version":1}