{"id":"a78eb513-73ac-436e-b875-d71975f90a5d","arxiv_id":"2501.03985","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In isolated dwarf spheroidal simulations, intermediate-mass black holes grow little but their AGN feedback measurably alters star formation and gas loss.","lead":"This paper simulates a galaxy like Leo II with a black hole at its center to see how its feedback affects star formation over 13.7 billion years. It finds the black hole rarely grows, but its energy output can suppress or even boost star formation under some conditions.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'tens of percent' AGN suppression for 10^4 Msun seeds is not robust: it hinges on the uncalibrated alpha=100 Bondi boost, is non-monotonic in alpha (Table 3), and is unconverged in resolution (Appendix).","rationale":"The paper's central claims are (1) IMBH seeds in isolated dSphs accrete little, and (2) AGN feedback nonetheless has non-negligible effects. Reading the simulations in good faith, claim (1) is supported by Fig. 4 and the parameter exploration; even with alpha=100, MBH,final/MBH,seed <2.4 in all runs. Claim (2) is qualitatively supported for 10^5-10^6 seeds, where suppression is severe even at alpha=1, but the specific 'tens of percent' statement for 10^4 seeds is the part that carries the paper's novelty and is the least secure. The authors themselves flag alpha as uncertain in Sec. 2.2, and Table 3's non-monotonic alpha dependence plus the Appendix's lack of strict convergence make the quantitative magnitude unreliable. I do not think the paper should be rejected; the qualitative conclusions and the honest parameter exploration justify conditional acceptance, exactly as the reader recommended. The proposed control runs with varied random seeds and higher resolution directly test whether Table 3 is reproducible. Agreement with the reader is 'agree' because the reader's weakest assumption already identified alpha and resolution as the key uncertainty; my analysis adds the specific non-monotonicity in Table 3 as concrete evidence that the quantitative suppression is not yet stable.","tokens_in":35065,"tokens_out":5221,"duration_ms":53543,"concrete_test":"Re-run TK4A100E1V3R, TK4A10E1V3R, and TK4A1E1V3R with five independent random seeds and at the Appendix's higher resolution (40,000 gas particles), holding all other parameters fixed. If the spread in Delta Mstar across seeds is larger than ~10 percentage points, or if the alpha=1 vs alpha=10 ordering reverses, then Table 3 does not establish a robust quantitative suppression for 10^4 seeds; if the ordering and magnitudes persist, the concern is settled.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline result that IMBH seeds grow little is robust across the parameter space (Fig. 4). The load-bearing problem is the quantitative companion claim that AGN feedback suppresses star formation by tens of percent for 10^4 Msun seeds. That magnitude comes from the fiducial alpha=100 Bondi boost (Eq. 1), which the authors state is uncertain for dwarfs (Sec. 2.2, Sec. 3.8). Table 3 shows the problem directly: for TK4A100E1V3R, Delta Mstar = -51%; for TK4A10E1V3R, Delta Mstar = -1%; for TK4A1E1V3R, Delta Mstar = -31%. The alpha=10 run contradicts 'tens of percent,' and the non-monotonic ordering (more suppression at alpha=1 than at alpha=10) is physically surprising, since lower alpha should reduce feedback energy. This indicates strong sensitivity or stochasticity rather than a robust relation. The Appendix adds that doubling resolution changes final stellar mass by +60% and shifts BH ejection times, while 'strict numerical convergence was not attained.' Thus the quantitative 10^4-seed feedback claim is conditional on alpha and resolution; the strong suppression for 10^5-10^6 seeds and limited BH growth are not in question.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents smoothed-particle hydrodynamic simulations of an isolated, Leo II-like dwarf spheroidal galaxy, extending the authors' earlier stellar-feedback-only model by adding AGN feedback from a central intermediate-mass black hole. It explores BH seeds from 10^3 to 10^6 M_sun and varies the Bondi boost factor, feedback efficiency, wind velocity, thermal versus kinetic feedback, BH repositioning, wind geometry, and radiative efficiency. The headline results are that the BH seeds grow very little (final-to-seed mass ratios remain below about 2.4), and that AGN feedback can suppress star formation substantially for 10^4-10^6 M_sun seeds, with some positive feedback for 10^3 M_sun seeds. The authors conclude that IMBHs in isolated dSphs grow inefficiently and that low-efficiency AGN feedback may still have non-negligible effects on the host galaxy.","tokens_in":35381,"tokens_out":3568,"duration_ms":35771,"significance":"If the quantitative feedback magnitudes were robust, the paper would provide a useful constraint on IMBH seeding mechanisms and on the calibration of subgrid AGN models in the low-mass galaxy regime. The paper's strengths include a broad parameter-space exploration, explicit reporting of simulation parameters in Table 2, and a dedicated resolution appendix. The robust part of the work is the limited BH growth, which is consistent across the explored parameter space. The less robust part is the magnitude of the AGN's effect on star formation, which depends strongly on the uncalibrated Bondi boost factor and on numerical resolution; the authors themselves acknowledge both uncertainties in Sections 2.2, 3.8, and the Appendix.","major_comments":[{"comment":"The claimed 'tens of percent' star-formation suppression for 10^4 M_sun seeds is not robust to the uncalibrated Bondi boost factor alpha. For runs TK4A100E1V3R, TK4A10E1V3R, and TK4A1E1V3R, the final stellar mass changes are -51%, -1%, and -31%, respectively, and the ordering is non-monotonic in alpha (alpha=1 suppresses more than alpha=10). Because Section 2.2 states that alpha=100 is uncertain for dwarfs and Section 3.8 reports that the Bondi radius is unresolved by factors of 7 to 100, the quantitative AGN-impact claim is conditional on alpha; the paper should either justify alpha for this regime or reframe the suppression magnitudes as upper/lower limits.","section":"Sec. 3.8, Table 3"},{"comment":"The resolution test shows that doubling the number of gas particles changes the final stellar mass by +60% (reduced to +20% when the stellar mass loading factor is retuned), shifts the BH ejection time to earlier cosmic times, and the text explicitly states that 'strict numerical convergence was not attained.' This directly affects the quantitative SFH comparisons in Section 3.2.1 and the BH ejection claim in Section 3.5; the paper should either demonstrate convergence for the reported magnitudes or present the AGN feedback effect as tentative and resolution-dependent.","section":"Appendix, Table 4 and Figs. 19-20"},{"comment":"All percentage changes in stellar mass are computed relative to a single fiducial stellar-only run from Hazenfratz et al. (2024), with no estimate of run-to-run stochasticity. Since the paper itself attributes one +1.3% case to stochastic fluctuations, the reader cannot distinguish physical sensitivity from numerical scatter in values ranging from -1% to -51%; reporting multiple stellar-only realizations or a stochasticity estimate would clarify which differences are meaningful.","section":"Sec. 3.2.1 and Fig. 6"}],"minor_comments":[{"comment":"The text 'around 600 Gyr' should read 'around 600 Myr' to be consistent with the time axis of Fig. 5.","section":"Sec. 3.2.1"},{"comment":"The word 'discrete' is repeatedly used where 'discreet' is meant, for example when describing the reduction in star formation activity and the attenuation of outflows.","section":"Secs. 3.2.1 and 3.3"},{"comment":"There is a typo in the discussion: 'negative feedback thay variably suppresses' should read 'negative feedback that variably suppresses.'","section":"Sec. 4"},{"comment":"Rows 3 and 10 both carry the label TK3A100E1V5 with identical parameters; this duplicate entry should be removed or renamed to avoid confusion.","section":"Table 2"},{"comment":"The caption states 'BHseed = 10^6 M⊙' but the text notes the yellow curve does not appear because star formation is completely suppressed; the caption should explicitly say that this curve is absent for that reason.","section":"Caption of Fig. 5"},{"comment":"The sentence describing the Bondi radius resolution, 'the spatial resolution being greater by 7 times up to 2 orders of magnitude than this parameter over time,' is awkward and should be rephrased for clarity.","section":"Sec. 3.8"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern lands: the limited-BH-growth conclusion is robust and interesting, but the quantitative claim of non-negligible AGN feedback on star formation is conditional on alpha and resolution, both of which the authors acknowledge. A revision that reframes the central claim around the robust limited-growth result and clearly marks the SF-suppression magnitudes as conditional would make the paper publishable. I would not reject it, but the current abstract and conclusions overstate the certainty of the feedback magnitudes."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a useful parameter-space study of AGN feedback in an isolated Leo II-like dwarf, and its robust result—that IMBH seeds barely grow—is credible. The softer claim that 10^4 solar-mass seeds suppress star formation by tens of percent is not robust; it depends strongly on the uncalibrated Bondi boost and on resolution, and the paper shows that honestly.\n\nWhat's new: they run 61 SPH simulations with thermal+kinetic AGN feedback from Barai et al. (2014) in a regime nobody had done before, seeds from 10^3 to 10^6 in a dSph with no mergers. They map the parameter space: alpha, epsilon_f, v_w, BH dynamics, wind geometry, radiative efficiency. The result that growth stays below about 2.4 times the seed across all runs is consistent and physically plausible—in a shallow potential, feedback self-regulates. The positive feedback at 10^3 solar masses and BH ejection without mergers are interesting, though both need confirmation.\n\nGood things: the paper is transparent. Table 3 shows the alpha sensitivity, the appendix shows the resolution test, and Section 3.8 says alpha is uncertain for dwarfs. Comparisons to Barai & de Gouveia Dal Pino (2019) and Koudmani et al. give context. The citation pattern is fine; self-citations are for the companion stellar-feedback paper.\n\nSoft spots, in order of severity: (1) The 'tens of percent' suppression for 10^4 seeds is load-bearing. Table 3 gives -51%, -1%, and -31% for alpha=100, 10, and 1. The alpha=10 run shows essentially no suppression, and the non-monotonicity suggests stochasticity rather than a robust relation. The abstract and summary lean on this magnitude. (2) Resolution: doubling particles changes final stellar mass by +60% and shifts BH ejection times; the paper admits no strict convergence. (3) The simulated stellar velocity dispersion reaches about 100 km/s, far above observed dSph values near 10 km/s, so the galaxy is not a perfect Leo II analog; this weakens quantitative comparisons. (4) BH wandering and ejection lack explicit dynamical friction; the authors themselves caution about this.\n\nWho it's for: people modeling AGN feedback in dwarfs and interpreting dwarf AGN observations. It gives a useful prior that isolated dSph IMBHs do not grow, and that suppression depends sharply on subgrid parameters. A serious referee can help tighten the quantitative claims, so I would send it to review.","headline":"Useful parameter-space study with a robust no-growth result for IMBH seeds in an isolated dSph, but the quantitative AGN suppression claims for 10^4 solar-mass seeds rest on shaky subgrid and resolution choices.","tokens_in":35949,"tokens_out":2362,"would_cite":true,"duration_ms":22958,"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 an isolated dwarf spheroidal galaxy, an intermediate-mass black hole barely grows but its feedback can still suppress or even boost star formation.","keywords":["dwarf spheroidal galaxies","AGN feedback","intermediate-mass black holes","hydrodynamical simulations","star formation history","black hole growth","Leo II","isolated galaxies"],"falsifier":"A high-resolution simulation that resolves the Bondi radius (or uses a calibrated subgrid model from such a run) for the same Leo II-like setup, checking whether the black hole grows beyond a factor of 2.4 and whether star formation is still suppressed by tens of percent; alternatively, an observational census finding a $10^{5}$ solar-mass IMBH in a dwarf spheroidal that is still forming stars would conflict with the paper's prediction that such a seed completely quenches the galaxy.","tokens_in":100,"feed_emoji":"🕳️","tokens_out":4229,"duration_ms":87323,"temperature":0.7,"pith_summary":"This paper asks whether an intermediate-mass black hole (IMBH) at the center of a dwarf spheroidal galaxy like Leo II can grow by gas accretion and whether its feedback meaningfully shapes the galaxy's evolution. Using isolated hydrodynamical simulations spanning 13.7 Gyr, it finds that the black hole seeds grow very little—never more than about 2.4 times their initial mass—regardless of seed mass, feedback efficiency, or wind velocity. Yet the same simulations show that AGN feedback is far from negligible: a $10^{4}$ solar-mass seed suppresses star formation by tens of percent, $10^{5}$ and $10^{6}$ solar-mass seeds quench it almost completely, and $10^{3}$ solar-mass seeds can even produce positive feedback that enhances star formation. The study matters because it suggests that isolated dwarfs with detected AGN could serve as clean laboratories for constraining black hole seeding mechanisms, and that low-luminosity AGN can regulate star formation in galaxies down to ~$10^{9}$ solar-mass halos.","feed_headline":"Dwarf black holes stay tiny yet still quench star formation","feed_subtitle":"Simulations show AGN feedback alters a Leo II-type galaxy even though the black hole barely grows.","key_machinery":"The argument is carried by a subgrid black hole accretion and feedback model built on the Bondi-Hoyle-Lyttleton accretion rate, corrected by a factor α (set to 100 in the fiducial runs) to account for the unresolved Bondi radius. The AGN feedback is implemented as thermal heating plus kinetic winds with an energy-driven outflow rate that depends on feedback efficiency εf and wind velocity vw. This machinery determines both how much the black hole grows and how much energy is injected into the surrounding gas, which in turn drives the star formation suppression or enhancement seen in the simulations.","core_discovery":"The central discovery is that IMBHs in an isolated dwarf spheroidal accrete so little gas that their final masses remain within the same order of magnitude as their seeds (MBH,final/MBH,seed < 2.4 across all tested models), while still exerting measurable feedback on the host galaxy. For a $10^{4}$ solar-mass seed with a 1% feedback efficiency, the final stellar mass drops by up to ~70% depending on wind velocity; at 5% efficiency the reduction is ~88-91%. Seeds of $10^{5}$ solar masses reduce stellar mass by 90-99%, and $10^{6}$ solar-mass seeds suppress star formation entirely. Conversely, $10^{3}$ solar-mass seeds with an intermediate wind velocity of 3000 km/s increase final stellar mass by ~8-11%, a case of positive AGN feedback attributed to shock compression of gas. The paper concludes that IMBHs probably need mergers or gas-rich interactions to grow, and that feedback efficiencies above 5% are likely too high for dwarf spheroidals like Leo II.","pith_inferences":["If isolated dwarfs do not grow their IMBHs, then the current occupation fraction in such galaxies directly reflects the seeding mechanism, turning local dwarf spheroidals into a practical test bed for distinguishing seed formation scenarios.","The positive feedback at vw = 3000 km/s suggests a non-monotonic relationship between AGN power and star formation; similar effects could appear in other low-mass galaxies if AGN-driven turbulence compresses gas rather than expelling it.","The strong dependence on the accretion factor α (Table 3) implies that resolving the Bondi radius, or calibrating α with high-resolution simulations, is essential for making quantitative predictions about AGN feedback in dwarfs.","The simulated ejection of some IMBHs from isolated dwarfs, if physically real, would produce free-floating black holes in the field, a prediction that could be tested with proper-motion surveys or gravitational-wave detections of IMBH mergers."],"forward_implications":["If a 10^4 solar-mass IMBH exists in a Leo II-type dwarf, it could be observationally hidden while still having altered the galaxy's star formation history by tens of percent, so the absence of AGN signatures does not rule out a significant feedback role.","The severe quenching caused by 10^5-10^6 solar-mass seeds implies that such massive IMBHs are unlikely to reside in dSphs that retain gas or show recent star formation.","The near-zero growth of isolated seeds suggests that observed IMBHs in dwarf galaxies must have grown through mergers or interactions, making occupied dwarfs a probe of past accretion history.","Feedback efficiencies calibrated on massive galaxies (εf ≥ 0.05) appear too aggressive for dSphs; lower efficiencies near 1% or below are more plausible in the low-mass regime.","Wandering black holes reduce feedback impact relative to a centrally fixed BH, so the dynamical state of the IMBH must be considered when interpreting feedback effects in real dwarfs."],"supporting_citations":[{"why":"Supplies the thermal and kinetic AGN feedback subgrid model (energy-driven winds) that this paper adopts and extends to dwarf spheroidals.","marker":"Barai et al. 2014"},{"why":"Provides the Bondi accretion prescription with the α = 100 resolution correction and the thermal feedback implementation that the simulations build on.","marker":"Springel et al. 2005"},{"why":"Defines the Bondi-Hoyle-Lyttleton accretion formula that determines black hole growth from local gas density, sound speed, and relative velocity.","marker":"Bondi 1952"},{"why":"Establishes the fiducial stellar-only feedback model and initial conditions for the Leo II-like galaxy that the AGN runs are compared against.","marker":"Hazenfratz et al. 2024"},{"why":"Provides cosmological IMBH accretion rates and feedback outcomes in more massive dwarfs that the paper compares to its isolated-galaxy results.","marker":"Barai & de Gouveia Dal Pino 2019"},{"why":"Supports the conclusion that low-efficiency AGN feedback can regulate star formation in low-mass galaxies, though calibrated for more massive dwarfs.","marker":"Koudmani et al. 2022"},{"why":"Documents wandering massive black holes in dwarfs and the low occupation fraction, motivating the dynamical tests and the interpretation of limited growth.","marker":"Bellovary et al. 2019"},{"why":"Constrains the halo mass and circular velocity adopted for the Leo II reference model in the simulations.","marker":"Strigari et al. 2007"}],"fun_headline_variants":["Black hole seeds stay tiny, still reshape dwarf galaxies","Dwarf black holes: barely any growth, real star formation impact","IMBHs in isolation: little accretion, notable feedback","Weak AGN, strong effect: dwarf galaxy evolution","Simulated Leo II: black hole barely grows, but feedback matters"],"cache_read_input_tokens":38016,"weakest_assumption_plain":"The Bondi accretion rate is multiplied by a factor α = 100 to compensate for the unresolved Bondi radius, and the fiducial AGN feedback parameters are taken from simulations of more massive galaxies; if the true α or coupling efficiency is much smaller, the magnitude of AGN feedback on the host weakens substantially, though the limited black hole growth is robust.","fun_headline_variants_meta":{"raw":{"variants":["Black hole seeds stay tiny, still reshape dwarf galaxies","Dwarf black holes: barely any growth, real star formation impact","IMBHs in isolation: little accretion, notable feedback","Weak AGN, strong effect: dwarf galaxy evolution","Simulated Leo II: black hole barely grows, but feedback matters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000614,"raw_usage":{"total_tokens":2896,"prompt_tokens":1029,"completion_tokens":1867,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":645,"completion_tokens_details":{"reasoning_tokens":1784}},"tokens_in":645,"tokens_out":1867,"duration_ms":12950,"temperature":1.0,"reasoning_tokens":1784,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:41:47.569071+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A high-resolution simulation that resolves the Bondi radius (or uses a calibrated subgrid model from such a run) for the same Leo II-like setup, checking whether the black hole grows beyond a factor of 2.4 and whether star formation is still suppressed by tens of percent; alternatively, an observational census finding a $10^{5}$ solar-mass IMBH in a dwarf spheroidal that is still forming stars would conflict with the paper's prediction that such a seed completely quenches the galaxy.","supporting_citations":[{"cited_title":"2005, Monthly Notices of the Royal Astronomical Society, 361, 776","cited_arxiv_id":null,"evidence_quote":"Provides the Bondi accretion prescription with the α = 100 resolution correction and the thermal feedback implementation that the simulations build on."},{"cited_title":"A., & Caproni, A","cited_arxiv_id":null,"evidence_quote":"Establishes the fiducial stellar-only feedback model and initial conditions for the Leo II-like galaxy that the AGN runs are compared against."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supports the conclusion that low-efficiency AGN feedback can regulate star formation in low-mass galaxies, though calibrated for more massive dwarfs."},{"cited_title":"E., Bullock, J","cited_arxiv_id":null,"evidence_quote":"Constrains the halo mass and circular velocity adopted for the Leo II reference model in the simulations."}],"review_version":1}