{"id":"96f9baa4-30fb-4e62-a4e3-ea3a16d60c48","arxiv_id":"2411.13666","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In the TNG100 simulation, high-mass galaxies quench inside-out through AGN feedback, while low-mass galaxies, especially satellites, quench outside-in through environmental gas removal.","lead":"This paper uses the TNG100 cosmological simulation to map how star formation is distributed inside galaxies, splitting galaxies by mass, whether they are central or satellite, and by properties like black hole mass and environment. It finds that massive galaxies shut down star formation from the inside out via black hole feedback, while low-mass galaxies are quenched from the outside in by their surroundings.","discovery_kind":"extension","skeptic_critique":null,"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript uses the TNG100 cosmological simulation to construct radial profiles of luminosity-weighted stellar age (age_L) and the resolved star-forming main-sequence offset (ΔΣ_SFR) for high-mass (M* > 10^10.5 Msun) and low-mass (M* < 10^10 Msun) central and satellite galaxies. For each population, the profiles are subdivided by intrinsic parameters (SMBH mass, cumulative AGN feedback energy, morphology) and environmental parameters (halo mass, local galaxy overdensity, satellite joining time, and change in star-forming gas mass). The central claim is that high-mass galaxies are quenched from the inside out by AGN feedback, while environmental processes, dominant for low-mass galaxies, quench galaxies from the outside in. In particular, low-mass satellite quenching is argued to occur primarily in massive host halos through gas depletion.","tokens_in":42019,"tokens_out":7954,"duration_ms":77605,"significance":"If confirmed, the paper provides a coherent and observationally relevant picture of quenching modes in a state-of-the-art cosmological simulation, linking AGN feedback to inside-out quenching at high mass and halo-driven processes to outside-in quenching at low mass. The analysis has notable strengths: bootstrap uncertainties, explicit and repeated caveats about simulation resolution and the unresolved-SFR floor, and an appendix (Appendix A) documenting inter-correlations among the parameters. The work is transparent and the claims are falsifiable. The main weaknesses are that the quantitative gradient metric is confined to R < 1Re and does not capture the outer-profile signatures on which the outside-in claim rests, the unresolved-SFR floor could contaminate the ΔΣ_SFR-based evidence, and the attribution to AGN feedback is partly degenerate with the tight stellar-mass–black-hole-mass correlation. These issues are addressable and do not, in my reading, invalidate the qualitative conclusions.","major_comments":[{"comment":"The fixed sSFR of 10^-12 yr^-1 assigned to bins with unresolved star formation is described in Section 2, and the text itself quotes McDonough et al. (2023) as noting that ΔΣ_SFR profiles are biased toward low values at intermediate radii. The abstract's claim that environmental processes drive quenching from the outside in is partly based on ΔΣ_SFR profiles at large radii (e.g., Fig. 8b for low-mass satellites in massive halos). The authors should demonstrate that this outer signal is not an artifact of the resolution floor, for instance by reporting the fraction of spatial bins at each radius that are assigned the fixed floor for each population, or by recomputing the profiles after excluding such bins.","section":"Section 2 and Fig. 2"},{"comment":"The gradient ∇1Re is defined as the slope of a line fit over 0 ≤ r ≤ 1Re (Section 2), and all quantitative summaries in Appendix B report this quantity. However, the outside-in quenching evidence is described as a positive or flat slope at R ≳ 1Re (e.g., §4.1 for high-mass centrals with 10^13.5 < MH < 10^14, and §4.2 for δ5 > 10^1.5). Because the reported gradient excludes the region where the outside-in signal appears, the summary figures do not quantitatively support the central claim. Add an outer-radius gradient over, say, 1–1.5Re, or an outer-to-inner age contrast, and include it in the appendix summaries.","section":"Section 2 and Appendix B"},{"comment":"The claim that AGN feedback drives inside-out quenching is based on the variation of profile shape and normalization with MBH and ΣEQM (Figs. 1–4). Appendix A (Fig. 17) demonstrates a tight correlation between stellar mass and MBH and ΣEQM for central galaxies, so the steepening gradients could be a mass effect. The text acknowledges these correlations but does not control for stellar mass. The authors should either show that the trends persist in narrow stellar-mass bins or perform a multivariate analysis (e.g., random forest or partial correlation) to separate the contributions of mass and BH properties. Without this, the assignment of the inside-out quenching to AGN feedback specifically remains underdetermined.","section":"Section 3.1, §5.1, and Appendix A"}],"minor_comments":[{"comment":"The word 'negligable' appears in the legends of Figures 3 and 4; it should be 'negligible'.","section":"Section 3.1 (Figs. 3 and 4 legends)"},{"comment":"The sentence 'This indicating that mass rather than morphology likely drives quenching in high-mass TNG100 galaxies.' should read 'This indicates that mass rather than morphology likely drives quenching...'","section":"Section 5.1"},{"comment":"The phrase 'a combination of stellar particles formed in the last 20 and 100 Myr' is ambiguous about whether the ΣSFR maps are constructed from two separate tracer populations that are combined or averaged; please clarify.","section":"Section 2"},{"comment":"The grey points that mark all individual low-mass central galaxies are difficult to identify behind the magenta density contours; a separate panel or a higher-contrast plotting style would improve readability.","section":"Figure 15"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of ApJ and presents a useful, systematic simulation-based study. The reader's concern about the unresolved sSFR floor is legitimate, but the age_L profiles provide independent support for the outside-in picture, so the issue is fixable with a robustness test rather than fatal. The mass–black-hole degeneracy is the more substantial concern and may require additional analysis. Overall, the paper is publishable after major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a thorough, well-documented extension of McDonough et al. (2023), giving a broad multi-parameter portrait of what shapes radial star-formation profiles in TNG100. The headline result—massive galaxies quench inside-out in step with BH mass and cumulative AGN feedback, low-mass galaxies quench outside-in through environment—is not a huge surprise, but here it is backed by a consistent set of profile decompositions across BH mass, feedback energy, morphology, halo mass, overdensity, joining redshift, and star-forming gas loss. That systematic mapping is genuinely useful for anyone comparing simulations to MaNGA/CALIFA.\n\nThe paper does several things well. It uses age_L profiles alongside ΔΣ_SFR, which matters because age_L is less affected by the unresolved SFR floor. It includes bootstrap errors, discusses parameter inter-correlations in Appendix A, and is unusually candid about resolution limits and the artificial sSFR floor. The appendix summaries of gradients and normalizations are a nice reference. I credit the authors for not hiding the messy bits.\n\nThe soft spot is the causal language. Statements like “driven by AGN feedback” and “dominant quenching process” go beyond what population-averaged profiles establish. BH mass and feedback energy are tightly correlated with stellar mass and morphology, and while the paper acknowledges this, it does not really break the degeneracy. The reader's worry about the sSFR floor is legitimate for the low-ΔΣ_SFR outskirts, but it is not fatal: the age profiles independently show the same outside-in signatures, and the paper explicitly flags the known bias. Still, the authors should quantify how much of the outside-in ΔΣ_SFR signal survives if unresolved bins are handled differently, and soften the causal verbs in the abstract and discussion.\n\nMinor issues: some key bins are small—40 quenched low-mass centrals is thin for a claim about that population—and the gradients are only measured to 1 R_e, which may miss the outer regions where outside-in quenching shows up. Neither undermines the main conclusions, but they do temper the strong claims.\n\nWho is this for? Anyone working on resolved star formation in simulations or linking TNG to IFS surveys. It is a reference-class descriptive paper, not a paradigm shift. It deserves a serious referee; the analysis is careful and reproducible, and the limitations are mostly stated. I would send it to review with a request for softened causal claims and a robustness check on the unresolved floor.","headline":"A careful, useful TNG100 profile atlas that maps intrinsic and environmental drivers of radial star-formation structure; the causal language outruns the correlation-based evidence, but the descriptive core is solid.","tokens_in":42451,"tokens_out":1756,"would_cite":true,"duration_ms":42471,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Using the TNG100 cosmological simulation, this paper argues that high-mass galaxies quench from the inside out through black-hole feedback, while low-mass galaxies quench from the outside in through their environments.","keywords":["galaxy quenching","inside-out quenching","outside-in quenching","AGN feedback","satellite galaxies","TNG100 simulation","radial star formation profiles","star-forming gas stripping"],"falsifier":"Recompute the low-mass satellite $\\Delta\\Sigma_{\\mathrm{SFR}}$ and $\\mathrm{age}_L$ profiles after removing or lowering the fixed star-formation floor, or repeat the same profile construction in a higher-resolution run of the same simulation; if the positive outer $\\mathrm{age}_L$ slopes and quenched outer bins weaken, the outside-in conclusion is driven by the resolution floor.","tokens_in":41849,"feed_emoji":"🌌","tokens_out":9595,"duration_ms":90511,"temperature":0.7,"pith_summary":"This paper uses the TNG100 cosmological simulation to ask what sets where star formation stops inside a galaxy. It builds radial profiles of luminosity-weighted age and of the offset from the resolved star-forming main sequence for thousands of central and satellite galaxies, split at $M_* = 10^{10.5}\\,M_\\odot$ and $10^{10}\\,M_\\odot$. The central claim is that high-mass galaxies, whether centrals or satellites, are quenched from the inside out by energy injected by their supermassive black holes, while low-mass galaxies are quenched from the outside in by their environments. Environmental effects on averaged profiles show up only at extreme halo masses and overdensities, and the strongest low-mass satellite signal is tied to how much star-forming gas a satellite has lost since joining its host halo. A sympathetic reader would care because this connects black-hole feedback and environmental gas stripping to measurable radial signatures of where star formation dies.","feed_headline":"TNG100: big galaxies quench inside-out, small ones outside-in","feed_subtitle":"Black-hole feedback sets radial shutdown in massive systems; host halos strip the outskirts of low-mass satellites.","key_machinery":"The load-bearing tool is the population-averaged radial profile. The authors project stellar particles of each galaxy face-on, bin by $R/R_e$, and compute the luminosity-weighted age $\\mathrm{age}_L$ from all particles in a bin and $\\Delta\\Sigma_{\\mathrm{SFR}}$, the logarithmic offset of the star-formation surface density from the resolved star-forming main sequence, with unresolved bins assigned a fixed floor of $\\mathrm{sSFR}=10^{-12}\\,\\mathrm{yr}^{-1}$. The gradient $\\nabla_{1R_e}$ measured from $0$ to $1\\,R_e$ turns a profile into one number: negative $\\mathrm{age}_L$ slopes and positive $\\Delta\\Sigma_{\\mathrm{SFR}}$ slopes mean inside-out quenching, while flat or positive outer $\\mathrm{age}_L$ slopes mean outside-in quenching. For satellites, the joining redshift $z_j$ and the change in bound star-forming gas mass $\\Delta M_{\\mathrm{SFG}}$ since joining serve as direct diagnostics of how much environmental exposure a galaxy has experienced.","core_discovery":"The paper's discovery is a division of labour between two quenching directions in the TNG100 universe. For high-mass galaxies ($M_*>10^{10.5}\\,M_\\odot$), the radial profiles steepen as black hole mass and cumulative quasar-mode feedback energy increase: gradients of $\\mathrm{age}_L$ become more negative and gradients of $\\Delta\\Sigma_{\\mathrm{SFR}}$ become more positive, signalling that star formation is being shut down first in the centre. Above roughly $M_{\\mathrm{BH}}\\sim10^{8.5}\\,M_\\odot$ and $\\Sigma E_{\\mathrm{QM}}\\sim10^{17}$ these galaxies are entirely quenched, and further feedback only flattens the age profile. For low-mass galaxies ($M_*<10^{10}\\,M_\\odot$) black hole properties barely change the profiles; instead, low-mass centrals that are quenched tend to lack a seeded black hole, and low-mass satellites are quenched when they live in halos above $M_H\\sim10^{13.5}{-}10^{14}\\,M_\\odot$, joined their hosts early, and have lost star-forming gas. The outskirts of such satellites are old and quenched while their centres can still form stars, which is the outside-in pattern. The overall conclusion is that intrinsic AGN feedback drives inside-out quenching at high mass, while environmental processes, primarily interactions with the host halo, drive outside-in quenching at low mass.","pith_inferences":["One extension the authors leave implicit is to apply the same $\\Delta M_{\\mathrm{SFG}}$ accounting to high-redshift satellites, which would test whether outside-in quenching begins only after group infall or can precede it.","If the fixed star-formation floor were removed or refined, the outer quenched regions of low-mass satellites should be re-examined; a weaker outside-in signal in a higher-resolution version of the simulation would indicate that part of the pattern is a resolution artefact.","The comparison with observed galaxies would be sharper if the profile construction mimicked the spaxel-binning used by integral-field surveys, since the paper notes that unresolved bins depress $\\Delta\\Sigma_{\\mathrm{SFR}}$ at intermediate radii."],"forward_implications":["High-mass galaxies in TNG100 quench through the same channel whether they are centrals or satellites, so black-hole feedback, not environment, sets their radial star-formation structure.","Completely quenched high-mass systems have old, flat $\\mathrm{age}_L$ profiles, so a flat profile should be read as the end stage of inside-out quenching rather than as a sign of environmental stripping.","Low-mass satellite quenching in TNG100 is controlled by the host halo: quenched systems cluster at $M_H>10^{13.5}\\,M_\\odot$ with early joining times and $\\Delta M_{\\mathrm{SFG}}<0$, even when local overdensity is modest.","Environmental quenching changes the outskirts first, so observations that measure beyond about one effective radius are needed to see it; inner gradients alone will miss the outside-in signal.","The spread within any environmental bin is large, meaning averages conceal a wide range of individual stripping histories."],"supporting_citations":[{"why":"supplies the sample selection, radial-profile construction method, resolved star-forming main sequence, and the known resolution bias that the new analysis inherits.","marker":"McDonough et al. 2023"},{"why":"provides the observed $\\Delta\\Sigma_{\\mathrm{SFR}}$ profiles and the thresholds in black-hole mass, halo mass, and overdensity that the simulation results are compared against.","marker":"Bluck et al. 2020b"},{"why":"establishes that kinetic AGN feedback in the TNG model reproduces central suppression of star formation and drives inside-out quenching.","marker":"Nelson et al. 2021"},{"why":"defines mass quenching and the stellar-mass threshold near $10^{10}\\,M_\\odot$ that the paper uses to separate quenching regimes.","marker":"Peng et al. 2010"},{"why":"supplies the environmental-quenching framework for satellite galaxies that motivates the halo-mass and overdensity analysis.","marker":"Peng et al. 2012"},{"why":"supports the conclusion that rapidly quenched galaxies are predominantly low-mass satellites, matching the low-mass satellite result here.","marker":"Corcho-Caballero et al. 2023"},{"why":"provides the synthetic-image morphological catalog from which the bulge statistic and Sérsic indices are drawn.","marker":"Rodriguez-Gomez et al. 2019"},{"why":"defines the Gini-M20 bulge statistic used to subdivide profiles by morphology.","marker":"Snyder et al. 2015"}],"fun_headline_variants":["TNG100: AGN quenches big galaxies inside-out, environment outside-in","Big galaxies die from the core, small ones from the edges in TNG100","Mass flips quenching direction: AGN inside, environment outside in TNG100","TNG100: Quenching direction scales with galaxy mass","AGN quenches big galaxies from core, environment from edge in TNG100"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion that low-mass galaxies quench from the outside in rests on the fixed star-formation floor of $10^{-12}\\,\\mathrm{yr}^{-1}$ assigned to spatial bins where TNG100 cannot resolve star formation, so the quenched outer regions could partly be a resolution effect rather than genuine environmental quenching.","fun_headline_variants_meta":{"raw":{"variants":["TNG100: AGN quenches big galaxies inside-out, environment outside-in","Big galaxies die from the core, small ones from the edges in TNG100","Mass flips quenching direction: AGN inside, environment outside in TNG100","TNG100: Quenching direction scales with galaxy mass","AGN quenches big galaxies from core, environment from edge in TNG100"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000716,"raw_usage":{"total_tokens":3290,"prompt_tokens":1091,"completion_tokens":2199,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":707,"completion_tokens_details":{"reasoning_tokens":2096}},"tokens_in":707,"tokens_out":2199,"duration_ms":14455,"temperature":1.0,"reasoning_tokens":2096,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:00:27.644764+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the low-mass satellite $\\Delta\\Sigma_{\\mathrm{SFR}}$ and $\\mathrm{age}_L$ profiles after removing or lowering the fixed star-formation floor, or repeat the same profile construction in a higher-resolution run of the same simulation; if the positive outer $\\mathrm{age}_L$ slopes and quenched outer bins weaken, the outside-in conclusion is driven by the resolution floor.","supporting_citations":[],"review_version":1}