{"id":"8fc8c73f-23f2-4679-82d2-44a9406c6aca","arxiv_id":"2510.26875","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"In 39 simulated isolated dwarfs, extended galaxies grow via merger-built stellar disks that form from high-angular-momentum gas-rich satellites.","lead":"Dwarf galaxies in these simulations start out small and round, and about half grow into large, thin, rotating disks over time. The growth is triggered by gas-rich satellite galaxies that spiral inward on aligned orbits, adding spin and supplying roughly a third of the cold gas that later forms stars.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Feedback calibration is the load-bearing uncertainty: Massive Dwarfs' weaker/less bursty feedback (admitted in §5.3.1) may be why stellar disks form at log M*<9; if real dwarfs are burstier, the disk-formation pathway could be rare rather than primary.","rationale":"The reader's weakest assumption and mine coincide: the sub-grid feedback model is the least secure external condition on which the central claim rests. All of the paper's internal evidence (disk–extended overlap, gas AM transfer, satellite gas fractions, Appendix B) is consistent and well presented, but it is all generated within one feedback implementation. The paper itself flags in §5.3.1 that its disks may form at lower masses because feedback is weaker/less bursty than FIRE-2, and Fig. 11 shows the whole sample is more concentrated than FIRE-2 dwarfs; concentration is independently identified as a disk-formation criterion by Hopkins et al. (2023). Thus the proposed high-AM-merger pathway may be a real but rare channel (m11b is a nice external echo), not the primary pathway. This is a correctness risk, not an internal inconsistency, and it cannot be resolved by the current paper's analysis. I also note the reader's secondary concern about the z=0 alignment statistic in Eq. (5): projecting past satellite orbits onto the final stellar AM axis mechanically favors disk galaxies. That is a real issue for the strength of the merger-alignment correlation, but it is secondary because Figure 6 and Appendix B provide time-resolved support; a flawed alignment statistic would weaken one piece of population-level evidence, whereas a flawed feedback model would invalidate the entire sample. No code or data are released, so independent re-analysis is currently impossible. The appropriate verdict remains CONDITIONAL: the claim is plausible and well-illustrated, but hinges on a single feedback implementation whose realism is explicitly uncertain.","tokens_in":31267,"tokens_out":8706,"duration_ms":90006,"concrete_test":"Select a matched subset of Massive Dwarfs initial conditions (e.g., r492, r968, r642, plus 3–4 compact/not-disk systems) and rerun them in GIZMO with FIRE-2 feedback at matched resolution, then measure the z=0 disk fraction, R_e distribution, and whether high-AM gas-rich mergers produce long-lived stellar disks. If disks below log M*≈9 become rare or the extended population is produced by feedback-driven expansion rather than by disks, the central claim is not robust. A cheaper observational complement: obtain integral-field stellar kinematics for a volume-complete sample of isolated dwarfs at 8<log M*<9 and test the paper's prediction that nearly all extended dwarfs (R_e>2 kpc) are rotation-supported disks.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim depends on the ChaNGa/Keller et al. (2014) superbubble feedback faithfully matching the burstiness and energy coupling of real dwarf ISM. The paper itself states (§5.3.1) that weaker or less bursty feedback may be why disks form at lower masses in Massive Dwarfs than in FIRE-2, and Fig. 11 shows both Massive Dwarfs sub-samples are more centrally concentrated than FIRE-2 dwarfs. Since Hopkins et al. (2023) identify a concentrated mass profile as the key criterion for disk formation, the disk-forming high-AM merger route identified here may be enabled by a subgrid recipe that creates deeper potential wells at log M*<9. If the real ISM is closer to FIRE-2, gas-rich mergers could still occasionally deliver angular momentum and build a disk (as in the m11b case discussed in §5.3.1), but the proposed pathway would not be primary: most dwarfs would remain dispersion-dominated, and extended dwarfs would more often arise from feedback-driven expansion (El-Badry et al. 2016; Chan et al. 2018). In that case the paper's 54% extended and 64% disk fractions would be properties of the feedback model, not of dwarf galaxies. The authors' honest admission does not reduce the risk; it makes the key branch of the argument explicitly dependent on one subgrid choice.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper analyzes 39 isolated dwarf galaxies (7.5 < log M*/M_sun < 9.1) from the Marvelous Massive Dwarfs zoom-in simulations to study the physical origin of extended dwarf galaxies. The authors find that most extended dwarfs (Re > 2 kpc) are rotation-supported stellar disks that formed gradually through the accumulation of angular-momentum-supported stars. They argue that gas-rich mergers on high-angular-momentum orbits spin up the central gas, create a disk, and trigger sustained star formation that secularly grows the half-light radius. The paper also quantifies that about 30% of cold gas at z=0 originates from merging satellites. They propose that disk formation is the main pathway for size growth, in contrast to feedback-driven expansion.","tokens_in":31633,"tokens_out":5760,"duration_ms":55785,"significance":"If correct, this paper offers a compelling merger-driven alternative to the commonly invoked feedback-driven expansion as the main mechanism for producing extended dwarf galaxies. The study benefits from a relatively large sample of high-resolution cosmological zooms, a clean particle-level case study in Appendix B, and a quantitative inventory of satellite gas contributions. However, the central causal inference is weakened by (i) an alignment statistic that is partly circular because it uses the final stellar AM as a fixed basis, and (ii) an admitted dependence of the disk-forming route on the subgrid feedback model. The paper is honest about these caveats but does not currently mitigate them, so the overarching claim that disk formation is the primary pathway remains model-dependent.","major_comments":[{"comment":"The alignment statistic A_star·g is measured relative to the z=0 stellar angular-momentum vector of the main galaxy. Because a disk is an AM-supported structure, any merger that successfully builds a disk will, almost by construction, have its orbital AM aligned with the final stellar AM; galaxies that never form a disk have a stellar AM vector that is not a meaningful basis. The reported steep correlation between \\bar{L}_{*·g} and R_e (log \\bar{L}_{*·g} ∝ 1.99 log R_e) is therefore inflated by the definition of the statistic. I recommend redefining the alignment using a basis that does not use final information (e.g., the AM vector of the main galaxy's gas at the time just before each merger, or the satellite's orbital plane), or at least showing that the result is robust to using such a basis.","section":"§4.2.2, Eqs. (4)–(8)"},{"comment":"The paper acknowledges that weaker or less bursty feedback in Massive Dwarfs compared with FIRE-2 may be why disks form at lower stellar masses, and Fig. 11 shows that both disk and not-disk Massive Dwarfs are more concentrated than FIRE-2 dwarfs. Since Hopkins et al. (2023) identify central concentration as the key criterion for disk formation, the high-AM merger pathway described here may be enabled by the subgrid feedback choice. If real dwarfs have burstier feedback, most dwarfs might remain dispersion-dominated and the proposed pathway might be rare. The claim that disk formation is the 'primary' means by which isolated dwarfs become extended (Section 4, item 4 in Section 6) should either be qualified as specific to the Massive Dwarfs feedback model, or supported with a sensitivity test using a burstier feedback variant.","section":"§5.3.1, Fig. 11"},{"comment":"The paper's causal claim—that disk growth is the primary driver of size growth—rests on the observation that extended galaxies are almost all disks and that size grows secularly. However, the analysis does not quantify the fractional contribution of disk star formation relative to other processes that could redistribute existing stars or expand the galaxy (feedback-driven expansion, tidal stirring, etc.). A quantitative decomposition of the stellar mass growth in the disk component versus other components (e.g., bulge, halo) over time, similar to the particle-level but done for the full sample, would more directly support the 'primary pathway' claim. Without this, the conclusion is to some degree an interpretation of the correlation.","section":"§4, Fig. 3"}],"minor_comments":[{"comment":"The caption says axis ratios are measured at R_e, while the text in §4.1 says 'at 2R_e'. Please reconcile.","section":"Figure 4 caption"},{"comment":"Typo: '0.1−03 Gyr' should read '0.1–3 Gyr'.","section":"Section 5.1"},{"comment":"The phrase 'mergers with high orbital angular momentum satellites on high angular momentum (spiraling-in) orbits' is redundant; consider simplifying.","section":"Abstract"},{"comment":"The disk definition (median jz > 0.5) is a reasonable kinematic criterion, but the results should state how sensitive the disk fraction (64%) is to the chosen threshold.","section":"Section 2.6.2"},{"comment":"The terms 'half-light radius' and 'effective radius' are used interchangeably; please standardize.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper is scientifically interesting and the simulations are a valuable resource. The main concerns are not about technical execution but about the interpretation of the alignment statistic and the generality of the conclusions. I recommend major revision rather than rejection because the internal physical story is plausible and the authors are transparent about the limitations. However, if a revision cannot resolve the circularity of the statistic or provide a robustness test against burstier feedback, the 'primary pathway' claim should be significantly softened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid, honest simulation paper with a clean story — in the Massive Dwarfs suite, extended dwarfs are almost always disks, and those disks form after high-angular-momentum, gas-rich mergers. The mechanism is not brand new conceptually, but the quantitative application to log M*<9 isolated dwarfs is, and the paper will be a useful reference. The main caveats are a partly circular alignment statistic and a feedback model dependence that the authors themselves flag; neither kills the paper, but both should be pushed in review.\n\nWhat's new: the 54% extended / 64% disk fractions, the ~30% satellite cold gas fraction, the classification of disk evolutionary paths (extended, compact, disrupted), and the Appendix B particle-level example. That appendix is the strongest part — it shows in one well-chosen case that the merger reorients gas, star formation follows, and the half-light radius grows. The paper also does a good job comparing with FIRE, TNG50, and observations, and it is unusually candid about its own limitations, especially Section 5.3.1.\n\nSoft spots, in order: (1) The A_star·g statistic (Eqs. 4-8) is partly circular. It measures merger alignment with respect to the z=0 stellar AM vector. For a galaxy that ends up as a disk, a merger that built the disk will by construction be aligned with that vector; for a galaxy without a disk there is essentially no well-defined stellar AM axis, so the statistic will tend to show misalignment. That means Figure 7's strong correlation between alignment-weighted merger AM and size is somewhat self-fulfilling. The paper's other evidence (Figure 6, Appendix B) is not circular, so the mechanism survives, but the population-level statistic needs careful qualification. (2) The feedback dependence. The paper admits that the ChaNGa/Keller superbubble model may be weaker or less bursty than FIRE-2, and that this could explain why disks form at lower masses here. That is an honest and important caveat, but it means the 'primary pathway' claim is conditional on the feedback implementation. If real dwarfs are as bursty as FIRE-2 suggests, the merger-built disk route might be a minority channel rather than the main one. This is a model-dependence limitation, not a calculational error. (3) Minor: one extended galaxy (r489) is not a disk, which is acknowledged and reasonable. No code or data are released, so independent replication is not yet possible — worth noting but common.\n\nOverall: this is a strong hypothesis paper. The internal evidence is clear, the limitations are stated, and the causal story is plausible, but the strength of the population-level claim is not fully warranted by the statistics as defined. A good referee will ask the authors to rerun the alignment analysis with, say, the gas AM at the time of the merger rather than the final stellar AM, and to quantify how sensitive the disk fractions are to feedback strength.\n\nWho should read it: people working on dwarf galaxy formation, morphology, and the UDG problem; observers interested in the size-mass relation below 10^9 solar masses. It deserves a serious referee — send it to review. I would cite it once the alignment statistic is either fixed or explicitly downgraded to a heuristic.","headline":"Solid, honest simulation paper; the merger-built disk mechanism is plausible and the internal evidence is strong, but the main alignment statistic is partly circular and the claim's reach depends on a feedback model the authors themselves suspect is weaker than FIRE.","tokens_in":32175,"tokens_out":3589,"would_cite":true,"duration_ms":71651,"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":"The paper argues that isolated dwarf galaxies grow into extended, large-radius systems not through bursty star-formation feedback puffing up their stars, but through the gradual build-up of rotation-supported stellar disks, and that this di","keywords":["dwarf galaxies","galaxy evolution","galaxy sizes","stellar disks","angular momentum","galaxy mergers","size-sSFR relation","cosmological simulations"],"falsifier":"A concrete test: measure the stellar specific angular momentum distribution of isolated dwarf galaxies with stellar masses around 10^8 to 10^9 solar masses. If most extended dwarfs are found to be pressure-supported rather than rotation-supported (median stellar j_z below 0.5), or if their gas disks are frequently misaligned with the stellar disk, the claim that disks are the main pathway to extended dwarfs would be falsified. Similarly, a simulation with a more bursty feedback model that, on the same initial conditions, fails to produce extended disks after high-angular-momentum mergers would","tokens_in":31138,"feed_emoji":"🌀","tokens_out":5506,"duration_ms":52521,"temperature":0.7,"pith_summary":"The paper argues that isolated dwarf galaxies grow into extended, large-radius systems not through bursty star-formation feedback puffing up their stars, but through the gradual build-up of rotation-supported stellar disks. Using high-resolution cosmological zoom-in simulations of 39 isolated dwarfs, it shows that these disks are triggered by gas-rich satellites on high-angular-momentum, spiraling-in orbits, which deposit angular momentum into the central galaxy's gas and reconfigure it into a disk. Star formation in that disk then accumulates stars on large, ordered orbits, secularly increasing the half-light radius beyond 2 kpc. This makes mergers, not feedback, the primary pathway to extended dwarf galaxies, and explains why some dwarfs remain compact.","feed_headline":"Mergers, not bursts, grow extended dwarf galaxies","feed_subtitle":"Simulations show extended dwarfs are disks grown by gas-rich, spiraling-in mergers, not feedback bursts.","key_machinery":"The central object is the specific angular momentum vector of gas and stars, quantified through a kinematic disk definition: a galaxy is a disk when the median z-component of its star particles' unit angular-momentum vectors exceeds 0.5. The key mechanism is angular-momentum transfer from an incoming gas-rich satellite on a non-radial, spiraling-in orbit, which 'stirs' the central gas into a disk. The paper introduces an alignment parameter, the cosine of the angle between the satellite's orbital angular momentum and the galaxy's present-day stellar angular momentum (set to zero for misaligned angles), to show that what matters is not the raw angular momentum of mergers but whether it is ali","core_discovery":"The central claim is that the formation of rotation-supported stellar disks is the main route by which isolated dwarf galaxies become extended in size. In the simulated sample, dwarf galaxies initially form compact (Re < 2 kpc). About half later experience a period of gradual size growth at relatively stable specific star-formation rate, becoming 'extended.' The paper identifies the mechanism as gas-rich mergers on high-angular-momentum (spiraling-in) orbits: the satellite's gas deposits angular momentum into the central gas, which settles into a disk aligned with the orbital plane; star formation then builds up a population of high-angular-momentum stars that increase the half-light radius.","pith_inferences":["If correct, the morphology of isolated dwarfs is set primarily by the orbital angular momentum of their past mergers; the dwarf size-mass relation could be used as a statistical probe of merger orbital configurations in cold dark matter.","The authors note that their feedback model is less bursty than some other simulations; a natural test is to rerun the same initial conditions with a more bursty feedback prescription. If the extended disks dissolve, the claim's empirical reach is confined to feedback models of this type.","The mechanism predicts that young stars in extended dwarfs should be strongly concentrated in a disk plane, while older stars remain in a roughly spheroidal distribution; resolved stellar-population studies of nearby dwarfs could check this directly.","A corollary not pursued in the paper: the same angular-momentum transfer process may also explain the formation of some ultra-diffuse dwarfs in isolation, if they are simply extended disk dwarfs seen face-on, rather than products of tidal stirring or feedback expansion."],"forward_implications":["Extended dwarf galaxies at z=0 should predominantly host rotation-supported stellar disks, with gas and young stars aligned in the same plane.","The scatter in the dwarf size-mass relation is largely a record of merger configuration and gas supply, not just stellar mass or star-formation burstiness.","Gas-rich, high-angular-momentum mergers can create long-lived stellar disks even in low-mass halos (down to about 10^7.7 solar masses), setting a lower mass limit for disk formation that depends on feedback physics.","Misaligned major mergers at late times can shrink an extended dwarf back to compactness, implying a transient population of high-mass, compact, non-disk dwarfs.","If the pathway holds, extended disk dwarfs should show a characteristic two-population stellar structure: an old spheroid plus a young, high-angular-momentum disk built over several gigayears."],"fun_headline_variants":["Dwarf galaxy size growth linked to high-spin mergers","Extended dwarfs are disks grown by spiraling-in mergers","Mergers on spiral orbits grow dwarf galaxy disks","Gas-rich mergers seed extended dwarf disk growth","How dwarf galaxies get big: disks from merger spin"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The paper's conclusions rest on the assumed sub-grid feedback model being realistic enough that supernova feedback is not so bursty that it prevents low-mass galaxies from ever settling into disks; if real feedback is more bursty, the merger-driven disk pathway could be rare or absent in actual dwarfs.","fun_headline_variants_meta":{"raw":{"variants":["Dwarf galaxy size growth linked to high-spin mergers","Extended dwarfs are disks grown by spiraling-in mergers","Mergers on spiral orbits grow dwarf galaxy disks","Gas-rich mergers seed extended dwarf disk growth","How dwarf galaxies get big: disks from merger spin"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000198,"raw_usage":{"total_tokens":1201,"prompt_tokens":740,"completion_tokens":461,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":385}},"tokens_in":484,"tokens_out":461,"duration_ms":12325,"temperature":1.0,"reasoning_tokens":385,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T07:03:18.137253+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test: measure the stellar specific angular momentum distribution of isolated dwarf galaxies with stellar masses around 10^8 to 10^9 solar masses. If most extended dwarfs are found to be pressure-supported rather than rotation-supported (median stellar j_z below 0.5), or if their gas disks are frequently misaligned with the stellar disk, the claim that disks are the main pathway to extended dwarfs would be falsified. Similarly, a simulation with a more bursty feedback model that, on the same initial conditions, fails to produce extended disks after high-angular-momentum mergers would","supporting_citations":[],"review_version":1}