{"id":"6d6a6583-6527-4c43-9e9c-8b9d3c5a7e49","arxiv_id":"2606.19463","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Simulations show observationally selected protocluster candidates at z ≳ 5 include significant interlopers, undergo 2-6 major mergers, and exhibit stronger clustering than observed, requiring total galaxy mass within 10 cMpc for reliable progenitor identification.","lead":"Using cosmological simulations, the study tracks the future evolution of galaxy groups identified as protocluster candidates at redshifts above 5. This reveals that many observed candidates include interlopers and will merge with other systems, suggesting better identification methods are needed.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.3","headline":"Simulations overpredict clustering relative to observations, undermining the merger probabilities and interloper fraction inferred from number density comparison.","rationale":"The reader's weakest assumption directly flags the simulation's reproduction of clustering and merger rates; the abstract's own reported discrepancy makes this the most load-bearing internal risk for the evolutionary interpretation, independent of selection details.","tokens_in":1860,"tokens_out":322,"duration_ms":12326,"concrete_test":"Compute the projected correlation function or neighbor counts at 5-10 cMpc for the observationally selected sample in FLAMINGO at z≈5-6 and compare directly to the equivalent statistic in the real observational catalog; if the simulated excess clustering exceeds 30%, recompute the merger probabilities using the observed neighbor distribution and check whether the number-density ordering still supports the claimed interloper fraction.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that the observed number density lying between the mass-selected and abundance-matched samples implies a mix of genuine progenitors and interlopers. This inference depends on the simulation correctly predicting merger rates and fates from the measured clustering (e.g., 2-10 neighbors within 10 cMpc, >50% merger probability at 5 cMpc). The abstract explicitly states that observations show markedly weaker clustering than FLAMINGO, creating an internal mismatch: if the simulated clustering (and thus merger dynamics) does not match reality, the positioning of the observed density between the two samples does not reliably constrain the interloper fraction or the later evolution.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper uses the FLAMINGO simulations to follow the z=0 descendants of protocluster candidates at z≳5. Three selection methods are compared: mass-selected, abundance-matched, and observationally selected. The observed number density is reported to lie between the mass-selected and abundance-matched samples, implying a mixture of genuine progenitors and interlopers. The simulations show strong clustering (2–10 neighbors within 10 cMpc), yielding merger probabilities of ≳50% (≳30%) for a neighbor at 5 cMpc (10 cMpc), mostly at z≲2. Observations exhibit weaker clustering than the simulations; each candidate experiences 2–6 later major mergers, mostly with systems too small to be identified at selection. The paper concludes that mass and overdensity alone are insufficient and that a total mass sum within 10 cMpc is required for robust identification.","tokens_in":2005,"tokens_out":497,"duration_ms":16140,"significance":"If the simulation-based merger statistics and selection comparisons hold, the work supplies quantitative guidance on the reliability of current high-z protocluster searches and identifies clustering amplitude as a potentially useful diagnostic. The large-volume simulation approach that tracks individual systems across cosmic time is a clear methodological strength.","major_comments":[{"comment":"Abstract: The inference that the observed number density lying between the mass-selected and abundance-matched samples implies a substantial interloper fraction rests on the assumption that FLAMINGO correctly predicts merger probabilities from the measured clustering (e.g., neighbor at 5 cMpc → ≳50% merger probability). The same abstract states that observations show markedly weaker clustering than the simulations, creating an internal tension: if the simulated clustering (and therefore the derived merger dynamics) does not match reality, the positioning of the observed density does not reliably constrain the interloper fraction or later evolution.","section":"Abstract"},{"comment":"Abstract: No numerical values, error bars, or tables are provided for the number densities of the three samples or for the observed sample, nor are the precise definitions of the mass-selected and abundance-matched selections given. Without these, the claim that the observed density 'falls between' the two simulated samples cannot be evaluated for statistical significance or sensitivity to parameter choices.","section":"Abstract"}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for their constructive comments, which help clarify the presentation of our results. We address each major comment below.","responses":[{"response":"The number density comparison and resulting implication for interlopers is based on the high-redshift selection statistics in the simulations versus observations, not directly on the merger probabilities. The merger probabilities are derived separately from the clustering properties within the simulations. We do note the discrepancy in clustering strength between simulations and observations, which we interpret as suggesting that clustering amplitude could serve as an additional diagnostic. However, we agree that this discrepancy introduces some uncertainty when applying the simulated merger rates to observed systems. We will revise the abstract to explicitly separate these points and add a caveat regarding the applicability of the merger statistics to observations given the clustering difference.","revision_made":"partial","referee_comment":"[Abstract] Abstract: The inference that the observed number density lying between the mass-selected and abundance-matched samples implies a substantial interloper fraction rests on the assumption that FLAMINGO correctly predicts merger probabilities from the measured clustering (e.g., neighbor at 5 cMpc → ≳50% merger probability). The same abstract states that observations show markedly weaker clustering than the simulations, creating an internal tension: if the simulated clustering (and therefore the derived merger dynamics) does not match reality, the positioning of the observed density does not reliably constrain the interloper fraction or later evolution."},{"response":"We agree that including specific numerical values and clearer definitions in the abstract would improve the accessibility of the key result. The full paper provides the definitions of the mass-selected and abundance-matched samples, along with the number densities in the relevant sections. We will update the abstract to include approximate number density values (with uncertainties where applicable) and brief definitions of the selection methods.","revision_made":"yes","referee_comment":"[Abstract] Abstract: No numerical values, error bars, or tables are provided for the number densities of the three samples or for the observed sample, nor are the precise definitions of the mass-selected and abundance-matched selections given. Without these, the claim that the observed density 'falls between' the two simulated samples cannot be evaluated for statistical significance or sensitivity to parameter choices."}],"tokens_in":1644,"tokens_out":480,"duration_ms":17400,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main point is that this paper tracks protocluster candidates from z ≳ 5 in FLAMINGO simulations and concludes that observed samples contain both real progenitors and interlopers, with neighbor-based merger probabilities of ≳50% at 5 cMpc. It also recommends summing total mass within a fixed 10 cMpc radius instead of relying on local overdensity.\n\nWhat is new is the explicit tie between observed neighbor counts at 5-10 cMpc and later merger odds, plus the direct comparison of number densities across mass-selected, abundance-matched, and observational selections. The paper does a clear job showing that each candidate experiences 2-6 major mergers and that the merger count stabilizes beyond ~10 cMpc.\n\nThe soft spot is the clustering discrepancy the abstract itself flags: observations show markedly weaker clustering than the simulations. The merger probabilities and the inference that the observed density sits between the two simulated samples both rest on the simulated clustering and merger dynamics being realistic. If they are not, the probabilities do not transfer to the real candidates. Methods details, error bars, and how the selections were matched to data are not visible in the abstract, so it is hard to judge how robust the quantitative results are.\n\nThis is for researchers working on high-redshift structure formation and survey planning. It has enough concrete claims to deserve a serious referee, provided the clustering mismatch is addressed head-on.","headline":"The clustering mismatch between FLAMINGO and observations undercuts the merger probabilities and interloper claims for real protocluster candidates.","tokens_in":2462,"tokens_out":361,"would_cite":false,"duration_ms":16431,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Observationally selected protocluster candidates at z ≳ 5 include both genuine cluster progenitors and many interlopers that will not reach cluster mass by z=0.","keywords":["protoclusters","high-redshift galaxies","galaxy clusters","cosmological simulations","structure formation","merger histories","number density"],"falsifier":"Direct measurement of how many observed z ≳ 5 protocluster candidates actually end up as z=0 clusters of the expected mass, compared against the simulated fractions.","tokens_in":2760,"feed_emoji":"🌌","tokens_out":773,"duration_ms":18592,"temperature":0.7,"pith_summary":"The paper uses simulations to follow what happens to protocluster candidates identified at redshift 5 and above in the same way observers select them. It compares the number density of these candidates to two other ways of picking systems in the simulations: by mass and by abundance matching. The real observed density sits between those two, showing that current searches capture some systems that will grow into clusters and many that will not. The candidates turn out to be strongly clustered, with a neighbor within 5 comoving megaparsecs raising the chance of later merger above 50 percent. This means forecasts based only on a candidate's mass and local overdensity at high redshift carry large scatter.","feed_headline":"High-z protocluster candidates mix true progenitors with interlopers","feed_subtitle":"Observed number density lies between mass-selected and abundance-matched samples in simulations, so many will not form clusters.","key_machinery":"Comparison of three selection methods (mass-selected, abundance-matched, and observational) within the FLAMINGO simulations to track later evolution and merger histories from z ≳ 5 to z=0.","core_discovery":"Using the FLAMINGO simulations, observationally selected protocluster candidates at z ≳ 5 have a number density that falls between the mass-selected and abundance-matched samples, indicating that searches recover both genuine cluster progenitors and significant interlopers that will not reach cluster masses by z=0. These candidates are heavily clustered, with 2-10 neighbors within 10 cMpc, and a neighbor at 5 cMpc (10 cMpc) implies ≳50% (≳30%) probability of later merging into a larger system, mostly at z ≲ 2. Each candidate experiences 2-6 later major mergers, mostly with systems too small to be identified as massive at the selection epoch, while observations exhibit weaker clustering than","pith_inferences":["Clustering strength around candidates could serve as an additional diagnostic to separate true progenitors from interlopers.","Upcoming wide and deep surveys can test the prediction by measuring total enclosed mass rather than just bright galaxy counts.","The mismatch between observed and simulated clustering may point to either selection biases in real data or missing processes in the models."],"forward_implications":["A candidate with a neighbor at 5 cMpc faces ≳50% chance of merging into a larger system, mostly at z ≲ 2.","Merger counts converge beyond ~10 cMpc, marking a characteristic scale in structure formation.","Relying only on high-z mass and galaxy overdensity produces severe scatter and systematic error in predicting final fate.","Robust identification requires summing total galaxy mass down to faint limits inside a 10 cMpc radius."],"fun_headline_variants":["High-z protocluster candidates include many interlopers","Sims indicate most high-z candidates merge with small systems","Clustering at z>5 signals later mergers for protocluster candidates","Need total mass within 10 cMpc to identify true progenitors","Observed clustering of protoclusters weaker than in FLAMINGO"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The simulations correctly capture the clustering, merger rates, and mass growth of these systems from high redshift down to the present.","fun_headline_variants_meta":{"raw":{"variants":["High-z protocluster candidates include many interlopers","Sims indicate most high-z candidates merge with small systems","Clustering at z>5 signals later mergers for protocluster candidates","Need total mass within 10 cMpc to identify true progenitors","Observed clustering of protoclusters weaker than in FLAMINGO"]},"model":"grok-4.3","cost_usd":0.005216,"raw_usage":{"total_tokens":2634,"prompt_tokens":880,"num_sources_used":0,"completion_tokens":85,"cost_in_usd_ticks":52162000,"prompt_tokens_details":{"text_tokens":880,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1669,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":880,"tokens_out":85,"duration_ms":12815,"temperature":1.0,"reasoning_tokens":1669,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-26T20:02:37.188547+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Direct measurement of how many observed z ≳ 5 protocluster candidates actually end up as z=0 clusters of the expected mass, compared against the simulated fractions.","supporting_citations":[],"review_version":1}