{"id":"57ea6bb3-a2f3-46ab-8420-6421f4b86ed5","arxiv_id":"2505.10524","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Stellar outflows in AGN disks can create a head-wind structure whose gravity accelerates the star forward (anti-friction), driving outward migration and potentially trapping stellar-mass black holes at equilibrium radii.","lead":"This paper runs 3D shearing-box simulations of stars with winds or jets embedded in AGN disks, and finds that the outflow can reverse dynamical friction, pushing the star outward instead of inward. It matters because the effect could change where stars and stellar-mass black holes collect in AGN disks, with possible consequences for black hole mergers and gravitational wave sources.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Fiducial outward migration requires sustained super-Eddington mass loss that §5.1 itself flags as astrophysically implausible for 8 M⊙ stars.","rationale":"The reader identified the same weak point — the super-Eddington mass-loss assumption — but still recommended CONDITIONAL. That verdict is too generous. The paper's own §5.1 admits the outflow rates cannot be sustained by 8 M⊙ stars, and the MASS model, which tests a star massive enough to potentially host such outflows, yields inward migration. This combination does not merely add uncertainty; it means the only astrophysically motivated configurations considered do not actually exhibit the claimed outward migration. The authors would need to show a self-consistent stellar model and simulation pair — a physically motivated ˙m and vsrc for an 8–100 M⊙ star with accretion — that produces positive ⟨ay⟩. Without this, the central claim rests on an outflow that likely never forms. The paper does provide useful technical groundwork: it demonstrates anti-friction in a numerically controlled setting, provides a parameter scan, and acknowledges the caveats. However, the headline claim is not just conditional on uncertain parameters; it is undermined by the paper's own cited physics. REJECT may sound harsh for a paper that is transparent about limitations, but the concern is about the central claim, not the quality of the numerics or the honesty of the discussion. If the authors can supply a self-consistent outflow model and simulation pair, the verdict should be upgraded. The requested test is concrete and directly addresses the gap.","tokens_in":18304,"tokens_out":2093,"duration_ms":18356,"concrete_test":"Run a grid of stellar models (e.g., MESA or GIZMO/MESA coupled models) for 8–100 M⊙ stars embedded in AGN disk environments with accretion rates from 0 to 0.01 M⊙ yr^-1, and compute the resulting steady-state mass-loss rates ˙m and wind velocities. Then map these rates onto the simulation parameter space of Table 2. If no self-consistent stellar model yields positive ⟨ay⟩ (outward migration) at the corresponding ˙m and vsrc, the paper's central astrophysical claim fails. A cheaper, purely numerical test: rerun FID-anti (8 M⊙) at the lower ˙m values permitted by §5.1 and check whether the acceleration remains positive. Note that this test is a consistency check on the paper's own stated physical limits; the authors have already cited the suppression by accretion in Chen et al. (2024, 2025).","verdict_should_be":"REJECT","load_bearing_attack":"The central claim of §3.1/Table 2 is that the fiducial 8 M⊙ star (FID-anti) acquires a positive azimuthal acceleration, ⟨ay⟩ = 0.24 × 10^-8 km s^-2, implying outward migration at ⟨ṙcir⟩ ≈ 2.4 km s^-1, driven by the anti-friction head-wind structure. This result depends on the assumed outflow rate ˙m ≈ 3 × 10^-3 M⊙ yr^-1 (§2.3, Eq. 8), which is justified by super-Eddington mass loss. However, §5.1 explicitly concedes that (1) sustained super-Eddington mass loss requires either long-lived interior energy release (difficult to sustain) or opacity-bump continuum driving; (2) the opacity-bump mechanism requires masses above 40–60 M⊙ (Cheng et al. 2024); and (3) recent radiative-hydrodynamic models of ~50 M⊙ stars in AGN-like environments find the mass loss suppressed by accretion at ~0.01 M⊙ yr^-1 (Chen et al. 2024, 2025). Thus the fiducial 8 M⊙ example likely cannot supply the assumed outflow. Moreover, the MASS model (80 M⊙, vsrc = 2 × 10^8 cm s^-1), which is in the presumably viable mass regime, produces the opposite sign: ⟨ay⟩ = -0.30 × 10^-8 km s^-2, i.e., inward migration. The mechanism is therefore not merely missing a plausible progenitor; the one astrophysically viable configuration tested yields the opposite dynamical outcome. The outward-migration mechanism as presented is contingent on outflow rates that the authors themselves identify as not achievable for the fiducial stellar mass, and the higher-mass case that might achieve such rates does not exhibit anti-friction. The paper does not identify any parameter regime where a real star simultaneously satisfies the mass-loss requirement and exhibits outward migration. This is a load-bearing concern because it directly undercuts the abstract's claim that anti-friction enables AGN stars to acquire angular momentum as a general outcome, reducing the result to a proof-of-principle for outflows that may not exist in AGN disks.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents 3D local shearing-box hydrodynamic simulations, performed with the GPU code Kratos, of a star or stellar-mass black hole with outflows embedded in the outer regions of an AGN disk. It considers isotropic winds and jet-like outflows, and varies the radial pressure-gradient offset xp, the adiabatic index gamma, and the background accretion flow. The central claim, developed in §3.1 and summarized in Table 2, is that the fiducial 8 Msun star with an isotropic super-Eddington outflow develops a positive time-averaged azimuthal acceleration, <ay> = 0.24 x 10^-8 km s^-2, which through Eq. (10), rdot_cir ~ 2 ay / Omega0, implies outward migration at about 2.4 km s^-1. This is attributed to an anti-friction head-wind structure, in contrast to the inward migration found in the no-outflow control. A jet case (BH-jet-z) is used to argue that an sBH with a z-directed jet can be trapped at an equilibrium radius in the range 3808-4006 Rsch. The paper explicitly acknowledges several caveats in §5.1 and §5.3, including the difficulty of sustaining the assumed super-Eddington mass-loss rates and the simplified treatment of stellar structure and disk thermodynamics.","tokens_in":18726,"tokens_out":7406,"duration_ms":71588,"significance":"If the central simulation result were robust, the paper would establish a qualitatively new dynamical channel in AGN disks: outflow-driven anti-friction that reverses dynamical friction and promotes outward migration of embedded stars, with potential implications for sBH retention, binary formation, and GW progenitor scenarios. The paper has clear strengths: the local shearing-box model is well posed, the no-outflow control is a natural benchmark, the code is documented, the parameter choices are transparent, and the authors are explicit about modeling limitations. However, the astrophysical significance is currently limited by three interconnected gaps: the fiducial mass-loss rate is disfavored by the authors' own §5.1 discussion; the MASS model, which lies in the plausible high-mass range, yields the opposite sign of migration; and no convergence tests or error estimates are presented for the reported accelerations. The anti-friction mechanism itself is physically plausible in the idealized setup, but the paper as written does not yet demonstrate that the outward-migration effect operates for realistic AGN star parameters.","major_comments":[{"comment":"The fiducial model assumes mdot ~ 3 x 10^-3 Msun/yr for an 8 Msun star, justified by the super-Eddington scaling of Eq. (8). The authors themselves state in §5.1 that sustained super-Eddington mass loss is difficult, that continuum-driven rates above ~10^-3 Msun/yr require masses above 40-60 Msun, and that recent radiative-hydrodynamic simulations find such mass loss suppressed by accretion at ~0.01 Msun/yr (Chen et al. 2024, 2025). The FID-anti run therefore computes positive <ay> for a parameter combination that §5.1 indicates is not realized by an 8 Msun AGN star. Because the outward-migration claim rests on this run, the manuscript needs either a revised fiducial setup using a viable outflow rate for 8 Msun, or an explicit reframing of the result as an idealized proof-of-concept whose domain of applicability is separately established.","section":"§2.3, Eq. (8); Table 1; §5.1"},{"comment":"The MASS-anti run, with an 80 Msun star and vsrc = 2 x 10^8 cm/s, is the one model whose mass lies in the range that could plausibly sustain a strong outflow, yet it gives <ay> = -0.30 x 10^-8 km s^-2, i.e., inward migration. This is not a small quantitative change; it reverses the sign of the central effect. Taken together with the fiducial result, the sign of migration is not robust across the tested parameter space, and the paper currently provides no demonstrated case of anti-friction-driven outward migration for a progenitor that can realistically supply the assumed outflow. I ask the authors to map the sign of <ay> in the (mdot, vsrc, mass) parameter space and to identify a region with outward migration that is compatible with viable mass-loss rates.","section":"Table 2; §3.1, MASS model"},{"comment":"The sBH trapping claim rests on the BH-jet-z simulation, but its runtime is reported inconsistently: Table 2 footnotes give tevo = 0.1 P, while Table 3 states tevo = 3 P. A duration of 0.1 orbital periods is too short to define a converged time-averaged acceleration or to support a migration picture. In addition, the equilibrium trap range 3808-4006 Rsch in Figure 15 is inferred from a sparse scan (roughly seven values of r0) with no error bars, no resolution study, and no discussion of how the averaging interval or interpolation affects the zero crossings. The trapped-zone conclusion is therefore not yet supported by the presented evidence.","section":"§4; Table 2; Table 3; Figure 15"},{"comment":"The FID-fric control, which sets the baseline sign of dynamical friction, is run in a smaller and differently shaped box (Lx,Ly,Lz = 50,100,150 AU) than the FID-anti model (75,150,75 AU), even though the text says all other parameters are unchanged. The friction/anti-friction comparison is therefore not strictly controlled, and no resolution or box-size convergence study is presented for any model. Moreover, the tabulated <ay> values are single numbers without uncertainties; the time series in Figure 2 show substantial fluctuations around the means, and without error bars or convergence tests it is not possible to assess whether a reported positive or negative <ay> of order 10^-9 km s^-2 is statistically significant.","section":"§3.1; Table 1; Figure 5; Figure 2"}],"minor_comments":[{"comment":"The displayed definition of xp appears corrupted: it reads 'xp = -∂xp/(2qΩ0^2ρ)' which is dimensionally inconsistent; it should presumably involve the radial pressure gradient ∂p/∂x.","section":"Eq. (6)"},{"comment":"The text states that the time-averaged acceleration is '≈ 0.40 km s^-2', but Table 2 uses units of 10^-8 km s^-2; please correct the units for consistency.","section":"§4"},{"comment":"The footnote legend using '+' and '-' to denote anti-friction and friction cases is not defined in the table header; please add a clear legend.","section":"Table 2"},{"comment":"The caption says 'Both models consider the scenario with stellar outflow', but the figure shows three JET models; please correct the wording.","section":"Fig. 11 caption"},{"comment":"The heating ratio η uses R0 before R0 has been defined; please define the standoff radius before or within Eq. (7).","section":"Eq. (7)"},{"comment":"There are several typographical issues, including 'there are several several caveats' and 'procedcures'; a careful proofreading pass is needed.","section":"Global"}],"recommendation":"major_revision","confidential_remarks":"My concern is not the anti-friction mechanism itself, which is well posed and physically plausible in the idealized simulation, but the bridge between the simulated parameters and realistic AGN star populations. The authors are transparent about the mass-loss caveats, and the framework is sound enough that I would not recommend rejection. However, the current fiducial example is arguably not realizable, the one high-mass example flips the sign, and the sBH trap claim lacks convergence and error analysis. A major revision that either demonstrates outward migration for a viable progenitor or clearly delimits the mechanism's parameter regime would substantially strengthen the paper."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear —, quick take on Liu, Wang & Peng (arXiv:2505.10524). What's actually new: applying the known anti-friction mechanism (Gruzinov et al. 2020; Li et al. 2020) to AGN-embedded stars, with a clean mapping of how the sign of the azimuthal acceleration depends on radial pressure gradient and outflow geometry. The jet case for sBHs—jet material confined to the trailing side, producing outward migration and a possible equilibrium trap—is a genuinely new suggestion and the most interesting part for the EMRI/merger community.\n\nThe paper does several things well. The fiducial runs are internally consistent: the head-wind structure is visible in the density maps, ⟨ay⟩ is positive, and Eq. (10) is the standard Gauss relation, so no free parameter is tuned to produce outward migration. The authors are also candid in §5.1 about the wind model's limits, which is more than many papers manage. The citation pattern is fine; the anti-friction lineage is credited.\n\nThe soft spots are real, though. The load-bearing assumption is the outflow rate. §5.1 concedes that ˙m ≈ 3×10^-3 M_sun/yr needs super-Eddington mass loss; for isolated stars that requires masses above 40–60 M_sun, and recent radiative-hydro simulations of ~50 M_sun stars in AGN-like environments find the mass loss suppressed by accretion. So the fiducial 8 M_sun case, the one that shows outward migration, is exactly the regime where the assumed outflow is implausible. And the one run in a plausible mass regime, MASS (80 M_sun), gives ⟨ay⟩ < 0—inward migration. That means the paper, as written, never demonstrates a parameter set in which a real star both satisfies the mass-loss requirement and migrates outward. That undercuts the abstract's general claim. It doesn't kill the mechanism as a proof of principle, but the claims need to be reframed.\n\nThe sBH trapping result is also thinner than the abstract implies. BH-jet-z runs only ~0.1–3 orbital periods, the equilibrium radius is inferred from a sparse scan over r0, and there are no error bars or convergence tests. The 'three balance points' in Figure 15 are read off a handful of simulations. No code or data are released, so others can't check quickly.\n\nWho is this for? People working on AGN disk migration, EMRI rates, or sBH assembly. I'd send it to a serious referee, but with a clear expectation of major revision: either add a viable progenitor model that shows outward migration, or reframe the paper as a proof-of-principle with explicit parameter requirements. The honest §5.1 suggests the authors know this.","headline":"A solid proof-of-principle for anti-friction in AGN disks whose central astrophysical assumption is admittedly shaky; the sBH trap result is too under-resolved to carry the abstract's weight.","tokens_in":19331,"tokens_out":2597,"would_cite":true,"duration_ms":25514,"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":"Stars and stellar-mass black holes with sufficiently strong outflows can gain angular momentum from AGN disk gas and migrate outward, reversing standard dynamical friction.","keywords":["stellar dynamics","active galactic nuclei","stellar winds","stellar mass loss","dynamical friction","anti-friction","stellar-mass black holes","shearing box simulations"],"falsifier":"Run the fiducial simulation with the mass-loss rate reduced by an order of magnitude (to about $3\\times10^{-4}\\,M_\\odot\\,\\mathrm{yr^{-1}}$) or with mass loss suppressed by disk accretion; if the time-averaged azimuthal acceleration is no longer positive, the outward migration and the equilibrium trap do not survive. A complementary observational check is to search AGN disks for an excess population of outflow-launching stars or black-hole merger sites at the predicted equilibrium radius.","tokens_in":18073,"feed_emoji":"🌬️","tokens_out":13205,"duration_ms":115663,"temperature":0.7,"pith_summary":"This paper argues from three-dimensional local shearing-box simulations that strong stellar outflows reverse the usual dynamical friction acting on objects embedded in active galactic nucleus (AGN) disks. In the fiducial case, an $8\\,M_\\odot$ star at about $5000$ Schwarzschild radii from a $10^8\\,M_\\odot$ black hole blows an isotropic wind, and the collision between that wind and the disk gas builds an overdense bow shock ahead of the star whose gravity accelerates the star forward instead of backward. The time-averaged azimuthal acceleration is $\\langle a_y\\rangle=0.24\\times10^{-8}\\,\\mathrm{km\\,s^{-2}}$, which through $\\dot{r}_{\\rm cir}\\approx 2a_y/\\Omega_0$ gives outward migration at $\\langle\\dot{r}_{\\rm cir}\\rangle\\approx 2.4\\,\\mathrm{km\\,s^{-1}}$; the same star without an outflow migrates inward. The effect is sensitive to the existence of a stable head-wind structure, which in turn depends on the outflow ram pressure balancing the ambient gas and on the disk's radial pressure gradient. A case study of a jet-launching stellar-mass black hole shows the same anti-friction can trap the black hole at an equilibrium radius, a zone where stellar-mass black holes might accumulate and merge.","feed_headline":"Outflows reverse drag on stars in AGN disks","feed_subtitle":"Simulations show a wind-driven bow shock pulls stars and black holes outward, not inward, in black-hole accretion disks.","key_machinery":"The load-bearing object is the head-wind (bow-shock) structure: the asymmetric density pattern formed where the stellar outflow's ram pressure meets the ambient gas streaming past the star. Its characteristic size is the standoff radius $R_0\\approx[\\dot{m} v_{\\rm src}/(4\\pi\\rho_0||\\mathbf{v}_g||^2)]^{1/2}$, the point where outflow and ambient ram pressures balance. The gravitational pull of the overdense bow shock on the star produces the positive azimuthal acceleration, and equation (10), $\\dot{r}_{\\rm cir}\\approx 2 a_y/\\Omega_0$, converts that acceleration into the orbital migration rate. Supporting machinery includes the local shearing-box hydrodynamics with a softened stellar potential and a spherical source region that continuously injects outflow, plus the pressure-gradient offset $x_p$ that sets the relative headwind speed between star and gas.","core_discovery":"The paper's central claim is that 'anti-friction' operates for outflowing stars in AGN disks: instead of a trailing overdense wake that drags the object back (ordinary dynamical friction), the star's wind creates a persistent overdense bow-shock structure ahead of the star, and the gravitational attraction of that structure accelerates the star forward. In the fiducial simulation, the star gains angular momentum from the disk gas at a time-averaged rate $\\langle a_y\\rangle=0.24\\times10^{-8}\\,\\mathrm{km\\,s^{-2}}$, which translates through $\\dot{r}_{\\rm cir}\\approx 2 a_y/\\Omega_0$ into outward migration at $\\langle\\dot{r}_{\\rm cir}\\rangle\\approx 2.4\\,\\mathrm{km\\,s^{-1}}$. The mechanism is sensitive to the radial pressure gradient of the disk: a steeper gradient (parameterized by $x_p=45$ AU) strengthens the outward acceleration, while a vanishing gradient ($x_p=0$) destroys the head-wind structure and restores inward migration. Isotropic winds are not the only route: in a case study of a stellar-mass black hole with a $z$-axis jet, the disk's high inflow velocity bends the jet material into the trailing side, and the resulting anti-friction gives $\\langle a_y\\rangle\\approx0.40\\times10^{-8}\\,\\mathrm{km\\,s^{-2}}$ and outward migration at $\\approx1.83\\,\\mathrm{km\\,s^{-1}}$. Varying the black hole's orbital radius reveals equilibrium points where inward and outward migration balance, trapping the black hole in a zone between roughly 3808 and 4006 Schwarzschild radii.","pith_inferences":["Population models of stars and stellar-mass black holes in AGN disks should include a two-sided migration torque: the main qualitative change is that outflow-launching objects can be parked at pressure-gradient-dependent radii rather than always drifting inward.","The equilibrium zone implies a spatial signature: gravitational-wave mergers of stellar-mass black holes born in AGN disks could cluster in an annulus of the disk rather than near the supermassive black hole.","The same head-wind anti-friction could in principle act on other outflow-launching bodies embedded in disks, such as massive planets in protoplanetary disks, though the much lower gas density makes the required ram-pressure balance harder to reach.","Because the simulations do not model turbulence, the stability of the head-wind structure in a realistically magnetized disk is open; adding turbulence is the most direct extension and could either disrupt the bow shock or add stochastic migration kicks."],"forward_implications":["Outflowing AGN stars can migrate outward instead of inward, so ordinary dynamical-friction capture is not the only possible fate for embedded stars.","The migration rate follows directly from the azimuthal acceleration through $\\dot{r}_{\\rm cir}\\approx 2a_y/\\Omega_0$; in the fiducial case it is $\\approx2.4\\,\\mathrm{km\\,s^{-1}}$, comparable in magnitude to inward frictional migration.","Anti-friction works only inside a parameter window: both too-weak and too-strong outflows fail (the $80\\,M_\\odot$ case shows inward migration), and the radial pressure gradient must be strong enough to maintain the head-wind structure.","Jet outflows along the $x$, $y$, or $z$ axes in the non-accreting fiducial disk do not produce anti-friction; the accreting-disk case with a $z$-axis jet is the one that yields outward migration.","An initially inward-migrating jet-launching stellar-mass black hole can settle in a trapped zone near $3808$–$4006$ Schwarzschild radii where inward and outward migration cancel, potentially accumulating black holes and fostering binary mergers."],"supporting_citations":[{"why":"Establishes the anti-friction effect: a sufficiently fast wind creates an underdense region around a moving object and reverses the gravitational drag.","marker":"Gruzinov et al. 2020"},{"why":"Supplies the hydrodynamic treatment and the standoff-radius scaling for how stellar outflows set the strength of anti-friction.","marker":"Li et al. 2020"},{"why":"Provides the AGN disk model, outer-region definition, and stellar properties (mass, radius, Eddington luminosity, surface temperature) used for the fiducial setup.","marker":"Cantiello et al. 2021"},{"why":"Provides the spherical source-region prescription for launching outflows and demonstrates positive torques from outflows in a binary context.","marker":"Wang & Li 2022"},{"why":"Supplies the shearing-box boundary conditions, softened-potential treatment, and the practice of running several orbits before measuring dynamical quantities.","marker":"Dong et al. 2011"},{"why":"Introduces the shearing-box approximation on which the entire local simulation method is based.","marker":"Hawley et al. 1995"},{"why":"Defines the pressure-gradient offset $x_p$ that parameterizes the headwind of gas relative to the embedded star and drives the anti-friction cases.","marker":"Masset 2017"},{"why":"Radiative hydrodynamic simulations finding that opacity-bump mass loss is suppressed by high accretion rates in AGN-like environments.","marker":"Chen et al. 2024"},{"why":"Extends the suppression result and is cited as evidence that super-Eddington mass loss at the modeled rates may not occur for embedded stars of about 50 solar masses.","marker":"Chen et al. 2025"},{"why":"Shows continuum-driven mass-loss rates above about $10^{-3}$ solar masses per year require stellar masses above roughly 40–60 solar masses, bounding when the mechanism can operate.","marker":"Cheng et al. 2024"}],"fun_headline_variants":["Wind bow shocks give AGN stars anti-friction outward drift","Anti-friction bow shock pulls AGN stars outward not inward","Outflow bow shocks make AGN stars migrate outward","Star winds create bow shock, reverse drag in AGN disks","Jets and winds push AGN stars outward via anti-friction"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The mechanism requires embedded stars to actually sustain strong outflows (mass-loss rates around $10^{-3}\\,M_\\odot\\,\\mathrm{yr^{-1}}$ or more), a condition the paper itself notes is hard to meet and may be suppressed by disk accretion.","fun_headline_variants_meta":{"raw":{"variants":["Wind bow shocks give AGN stars anti-friction outward drift","Anti-friction bow shock pulls AGN stars outward not inward","Outflow bow shocks make AGN stars migrate outward","Star winds create bow shock, reverse drag in AGN disks","Jets and winds push AGN stars outward via anti-friction"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000699,"raw_usage":{"total_tokens":3256,"prompt_tokens":1141,"completion_tokens":2115,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":757,"completion_tokens_details":{"reasoning_tokens":2030}},"tokens_in":757,"tokens_out":2115,"duration_ms":14636,"temperature":1.0,"reasoning_tokens":2030,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:08:17.411338+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the fiducial simulation with the mass-loss rate reduced by an order of magnitude (to about $3\\times10^{-4}\\,M_\\odot\\,\\mathrm{yr^{-1}}$) or with mass loss suppressed by disk accretion; if the time-averaged azimuthal acceleration is no longer positive, the outward migration and the equilibrium trap do not survive. A complementary observational check is to search AGN disks for an excess population of outflow-launching stars or black-hole merger sites at the predicted equilibrium radius.","supporting_citations":[{"cited_title":"2022, ApJ, 932, 108, doi: 10.3847/1538-4357/ac6ce6","cited_arxiv_id":null,"evidence_quote":"Provides the spherical source-region prescription for launching outflows and demonstrates positive torques from outflows in a binary context."}],"review_version":1}