{"id":"444b1849-8fcd-4fd7-8a39-1bb2d0cf8b9d","arxiv_id":"2611.17033","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":13,"one_line_summary":"NGC 985 hosts a recurrent, variable compact millimeter synchrotron component with a causality-based size limit near 1100 Schwarzschild radii.","lead":"Three ALMA observations of the Seyfert galaxy NGC 985 show a compact millimeter component that weakened sharply within one month and then partly recovered, placing a model-independent upper limit of about 0.024 pc on the emitting region. A smart generalist might care because variable compact millimeter emission is a candidate probe of the magnetized corona around supermassive black holes, which is also a possible site of neutrino production.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"Month-scale variability rests on the QA-failed Epoch-1a Band 4 execution; if its high flux traces an amplitude-calibration artifact rather than the source, the 10^3 R_S causality limit collapses.","rationale":"The paper's central claim is unusually well supported internally: the variability argument uses raw ALMA photometry rather than the coronal decomposition; the Epoch-1a and Epoch-1b beams are comparable in size, excluding a resolution artifact; Epoch-2 shows an intermediate compact detection; and every model-dependent inference (p, B, R, cooling times) is explicitly conditional. The residual soft spot is therefore observational rather than theoretical: the two ALMA Band 4 executions that define the month-scale change are not symmetric, because the Epoch-1a reduction failed delivery QA and was reobserved, and the cause is not disclosed. Since the abstract's headline is exactly the ~10^3 R_S causality limit, the amplitude scale of the Epoch-1a Band 4 data is the single most load-bearing datum in the paper. A factor-of-~2.5 epoch-band calibration error is far larger than the 5% correlated calibration term in the fit, so the standard error budget cannot rescue it if the QA problem is calibration-related. I am not asserting that the Epoch-1a measurement is wrong; I am recommending that unconditional acceptance be conditional on the one re-reduction/QA check that settles it. If the check passes, ACCEPT is appropriate; if not, the remaining Epoch-1b versus Epoch-2 comparison gives only a much weaker year-scale limit, and the central claim must be downgraded.","tokens_in":27027,"tokens_out":14319,"duration_ms":105665,"concrete_test":"Obtain the ALMA QA2/QA3 report and data-quality diagnostics for project 2015.1.01516.S, Epoch-1a Band 4 (2016-03-08) from the ARC, and independently re-reduce the raw ASDM with the contemporaneous J0238+1636 calibrator model using the current CASA pipeline; repeat the same re-reduction for Epoch-1b. If the independent Epoch-1a Band 4 flux is consistent with ~2.2 mJy within combined errors and the QA failure is documented as non-calibration (e.g., missing delivery metadata), the month-scale variability claim stands. If the re-reduction instead yields ~1 mJy or QA flags show anomalous antenna amplitude/gain solutions unique to Epoch-1a, the intrinsic month-scale variability and the R_var <= 0.024 pc limit collapse, and the verdict should be revised.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim (Section 4.2) is that the factor-~2.5 drop in Band 4 flux between Epoch-1a and Epoch-1b is intrinsic, giving R_var <= c Delta_t_obs/(1+z) ~ 0.024 pc ~ 1.1 x 10^3 R_S. Section 2.1 reports that the first Band 4 execution passed the initial QA0 assessment but the reduced products subsequently failed to meet the quality standards for delivery, so the scheduling block was reobserved on 2016 April 7; both executions were retained. The paper never states why the first reduction failed. If that failure corresponds to an amplitude-calibration, gain, or decorrelation problem specific to the Epoch-1a data, the high 2.2 mJy state could be instrumental, and the month-scale variability, and therefore the size limit, would not be established. Comparable beam sizes rule out a resolution artifact, and the Epoch-2 intermediate detection supports a compact component, but Epoch-2 is ~17 months later; without Epoch-1a the remaining Epoch-1b/Epoch-2 comparison gives only a year-scale limit. The 5% correlated calibration factor in the joint fit cannot absorb a factor-of-2 to 3 epoch-band offset. The J0238+1636 drift is not itself a clean explanation: an untracked decrease of the calibrator would bias the target flux upward, strengthening rather than removing the drop; the sharper systematic is the QA-failed Epoch-1a Band 4 reduction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper presents multi-epoch ALMA and VLA observations of the type-1 Seyfert galaxy NGC 985, reporting a variable compact millimeter excess. The central claim is that the millimeter flux density decreased by a factor ~2.5 between two ALMA epochs separated by about one month (Epoch-1a to Epoch-1b), and was detected again at an intermediate level in a third epoch ~17 months later (Epoch-2). This variability is argued to be intrinsic, implying a model-independent causality size limit of R_var ≲ c Δt_obs/(1+z) ≈ 0.024 pc ≈ 1.1×10^3 R_S. Spectral decomposition requires a compact, self-absorbed synchrotron component, and a hybrid thermal/nonthermal coronal model can reproduce the radio–millimeter spectra, with the two well-sampled epochs favoring different nonthermal electron distributions. The paper concludes that NGC 985 hosts a recurrent and variable compact millimeter synchrotron component, with a magnetized corona as a plausible origin, while not uniquely excluding alternatives such as a jet base.","tokens_in":27439,"tokens_out":1825,"duration_ms":13469,"significance":"If the variability claim holds, the paper provides one of the cleanest model-independent size constraints on the compact millimeter-emitting region in a radio-quiet AGN, using only causality from the observed flux change. This is significant because it establishes a compact nuclear origin without relying on spectral modeling. The multi-epoch spectral decomposition also allows comparison of spectral states within a common framework, which is relatively rare for such sources. The paper is careful to delineate model-dependent and model-independent results, and explicitly acknowledges limitations (e.g., the η–B–R degeneracy, the non-unique coronal interpretation). The availability of archival VLA data presented for the first time adds value. The central variability claim is supported by comparable beam sizes across the two key epochs, which rules out resolution artifacts. The paper's strength lies in the combination of variability and spectral constraints, and its honest discussion of degeneracies.","major_comments":[{"comment":"The central variability claim (Epoch-1a to Epoch-1b Band 4 flux drop) rests on the reliability of the Epoch-1a Band 4 data, which the paper itself reports as having failed QA after initial QA0 assessment (§2.1: 'the reduced products subsequently failed to meet the quality standards for delivery'). The paper states that both executions were retained, but never explains why the first reduction failed. If the failure was due to an amplitude-calibration or gain problem specific to Epoch-1a, the high 2.2 mJy flux could be instrumental, and the month-scale variability, and therefore the causality size limit, would not be established. The comparable beam sizes between Epoch-1a and Epoch-1b argue against a resolution artifact, but they do not address a calibration artifact. The paper should provide a specific explanation for the QA failure and demonstrate that it does not affect the amplitude calibration of the target. Without this, the model-independent size limit is on uncertain footing.","section":"§2.1, §4.2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The spectral decomposition assumes that the compact power-law, star-formation synchrotron, and dust components are common to both Epoch-1a and Epoch-2 (§3.4: 'assumed to be common to the two epochs'). This is a strong assumption, given that the paper argues for intrinsic variability of the compact component. If the star-formation synchrotron or dust emission also varied (e.g., due to different beam-filling factors or calibration offsets between epochs), the decomposition could be biased. The paper should justify this assumption or test its sensitivity, for example by allowing these components to vary within reasonable priors.","section":"§3.4, Table 3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The paper reports that 11.8% of Monte Carlo realizations for Epoch-2 reach the imposed lower bound p=2.02, and the 95% Monte Carlo interval of Δp = p_1a - p_2 includes zero. Despite this, the paper states that the two epochs favor different nonthermal electron distributions and that p_1a > p_2 in 92.1% of realizations. While the latter fraction is reported, the text should more prominently acknowledge that the difference in p is not statistically significant at the 95% level, given the large uncertainties and the boundary effect. The current phrasing ('favors a steeper nonthermal electron distribution') may overstate the significance.","section":"§4.3, Table 3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The paper demonstrates that the η–B–R degeneracy is strong, with B and R changing substantially when η is varied from 0.01 to 0.10 while the spectra remain nearly indistinguishable. This is correctly acknowledged. However, the paper does not discuss the possibility that η itself could vary between epochs, which would further complicate the interpretation of the epoch-to-epoch differences in B and R. Since the coronal parameters are the main physical output of the spectral fitting, this limitation should be explicitly stated.","section":"§4.4","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The paper states that the first Band 4 execution 'passed the initial QA0 assessment' but 'the reduced products subsequently failed to meet the quality standards for delivery.' It would be helpful to specify what aspect of the data failed (e.g., amplitude calibration, phase stability, decorrelation) and why the data were still considered usable.","section":"§2.1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The intrahour light curves show several >3σ excursions, but the paper correctly notes they are not secure evidence of intrinsic variability. The text could briefly note whether these excursions correlate with any known instrumental issues (e.g., atmospheric phase noise, pointing offsets) to strengthen this assessment.","section":"§3.2, Figure 2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The fitting objective uses a half-weighting for VLA data shared between the two epoch branches. The paper should justify this choice more explicitly, as it effectively reduces the constraining power of the VLA data. A brief sensitivity test showing that the results are not strongly dependent on this weighting would be reassuring.","section":"§3.4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The comparison with Swift/BAT hard-X-ray variability uses monthly binned data. The paper notes the limited significance and non-simultaneity. It might be worth adding a brief comment on whether any contemporaneous X-ray observations (e.g., Swift/XRT or NuSTAR) exist for the ALMA epochs, even if not published.","section":"§4.5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The figure caption refers to 'compact jet' in the legend, but the text uses 'compact power-law component' or 'compact jet-related radio emission.' The legend should be consistent with the text to avoid confusion.","section":"Figure 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The introduction cites several recent millimeter variability studies. It would be helpful to briefly note whether any of those studies also derived causality size limits, to contextualize the NGC 985 result.","section":"§1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The VLA data reduction uses different Perley-Butler models for different bands (2017 for L band, 2013 for C/X/K bands). The paper should note whether this introduces any systematic offset between bands, and if so, how it is accounted for in the spectral fitting.","section":"§2.3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The table lists flux densities for individual SPWs. It would be helpful to also list the weighted-mean flux density for each band and epoch, as this is what is typically shown in the figures.","section":"Table 1","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"The central result holds up: NGC 985's compact millimeter component dropped by a factor ~2.5 between two ALMA epochs a month apart and was detected again at an intermediate level 17 months later, and the resulting causality size limit of ~1.1x10^3 R_S is genuinely model-independent. You can trust the variability more than the corona, and the authors know it.\n\nWhat's new: two previously unpublished Cycle 3 ALMA epochs (2016 March and April), a careful reanalysis of the 2017 Cycle 4 epoch, and the first presentation of archival post-upgrade VLA data. The month-scale variability and the size limit are not in Inoue & Doi 2018. The beam-matched photometry is the right tool and is done well: Epoch-1a and 1b beams are comparable (~1.4-1.5 arcsec), so the ~1.3 mJy drop at 150 GHz is not a resolution artifact, and Epoch-2's much smaller beam still recovering an intermediate flux confirms a genuinely compact component.\n\nThe paper is unusually transparent. It flags the 222 GHz extended emission as only marginally detected (~2 sigma), downgrades the epoch-to-epoch electron-index difference to a tendency because the 95% Monte Carlo interval includes zero, admits the synchrotron turnover between VLA K band and ALMA is not directly sampled, and states plainly that the fitted B and R are conditional on the assumed eta = 0.04. Section 4.6's three-level summary (model-independent, decomposition-level, coronal-model-level) is exactly how this kind of result should be presented. Even the MCMC appendix admits its chains are too short for reliable posteriors.\n\nSoft spots, in proportion. First: the QA failure. Section 2.1 reports that the first Epoch-1a Band 4 execution passed QA0 but its reduced products failed delivery quality standards, so the block was reobserved; both executions were retained. The paper never says why the first reduction failed. Because the month-scale variability and the 10^3 R_S limit both rest on the Epoch-1a Band 4 flux, a referee should ask the authors to state the cause and show the absolute flux scale is unaffected. I would not call this fatal: the same-night Band 3 data show the same high state, the intrahour light curve is flat at that high level, and the J0238+1636 drift, if anything, would bias the Epoch-1b flux upward, so it cannot manufacture the drop. But the omission is the one real gap in an otherwise candid paper. Second: the abstract says the two epochs favor different nonthermal electron distributions, which mildly overstates a result the body explicitly downgrades to a tendency. Third: the coronal parameters are fitted, strongly covariant, and conditional; the authors say all of this themselves, so it limits the interpretation, not the paper.\n\nWho it is for: anyone working on radio-quiet AGN millimeter emission or coronal neutrino production. I would bring it to reading group, and it deserves a serious referee; the main request would be a full accounting of that Epoch-1a execution.","headline":"A transparent multi-epoch ALMA study whose month-scale variability and ~10^3 R_S causality limit for NGC 985's compact millimeter component hold up well, with one unresolved QA-execution question a referee should press on.","tokens_in":28113,"tokens_out":8731,"duration_ms":69525},"pith_extraction":{"msc":[],"pacs":["98.54.Cm"],"model":"deepseek/deepseek-v4-flash","headline":"Three ALMA epochs show the millimeter emission of the Seyfert galaxy NGC 985 collapsing by more than half within a month and reappearing at an intermediate level a year later, which limits the variable emitter to about 1,100…","keywords":["NGC 985","Seyfert galaxy","millimeter variability","AGN corona","synchrotron self-absorption","ALMA observations","radio-quiet AGN","supermassive black hole"],"falsifier":"Re-derive the Epoch-1a and Epoch-1b flux scales from an independent absolute calibrator or from ALMA's planet-based scale: if a $\\sim$1.3 mJy Band 4 difference between 2016 March 8 and April 7 survives, the intrinsic-variability claim and the $\\lesssim 0.024$ pc size limit stand; if it shrinks to the $\\sim$5% calibration floor, both collapse. A prospective check is weekly ALMA Band 4 monitoring of NGC 985 over several months: if the 150 GHz flux stays constant to within calibration noise, the month-scale drop was a singular event rather than a recurrent property, and the claim that NGC 985 hosts a recurrent variable component would need revision.","tokens_in":28362,"feed_emoji":"📡","tokens_out":20994,"duration_ms":128873,"temperature":0.7,"pith_summary":"The paper aims to establish that the compact millimeter excess of the radio-quiet Seyfert galaxy NGC 985 is a recurrent, intrinsically variable component rather than a stable spectral feature. Three ALMA epochs catch the 150 GHz flux collapsing from about 2.2 mJy to about 0.9 mJy in roughly one month and then reappearing at an intermediate level a year later; because the two bracketing epochs have comparable angular resolution, the drop itself bounds the emitting region to $\\lesssim 0.024$ pc, or roughly $1.1\\times10^3$ Schwarzschild radii (about a thousand times the black hole's event-horizon scale), without any spectral model. The paper then shows, with more model dependence, that the radio-to-millimeter spectrum requires a compact, self-absorbed synchrotron component, and that a magnetized black-hole corona with continuously reinjected nonthermal electrons reproduces the two well-sampled epochs with different electron distributions. If right, millimeter variability becomes a direct, assumption-light probe of the innermost region of a supermassive black hole, the same zone thought to produce the X-ray corona.","feed_headline":"Month-scale drop pins NGC 985's millimeter emission to 1,100 R_S","feed_subtitle":"No spectral model needed: the drop alone bounds the emitter to ~0.024 pc, a scale pointing to a black-hole corona.","key_machinery":"The machinery has two parts. The decisive one is the causality argument from variability: a flux change observed over a rest-frame interval $\\Delta t_{\\rm obs}/(1+z)$ cannot originate from a region larger than $c\\,\\Delta t_{\\rm obs}/(1+z)$, so the roughly 30-day Epoch-1a-to-Epoch-1b drop alone confines the variable emitter to $\\lesssim 0.024$ pc ($\\simeq 1.1\\times10^3 R_{\\rm S}$), with no spectral decomposition or coronal model required. The second is the multi-epoch spectral decomposition, which separates the compact nuclear component from extended star-formation synchrotron and thermal dust using the beam-size dependence of the measured flux densities, and describes the compact emission with a self-absorbed synchrotron spectrum from a hybrid thermal/nonthermal electron population, fitting the nonthermal index $p$, magnetic field $B$, and radius $R$ per epoch under a shared continuum. The authors stress that $p$, $B$, and $R$ are strongly covariant, with the quoted $B$ and $R$ values conditional on the fixed nonthermal energy fraction $\\eta=0.04$, whereas the causality limit is independent of all of these choices.","core_discovery":"On the paper's own terms, the discovery is that NGC 985 hosts a recurrent and variable compact millimeter synchrotron component. The evidence is three ALMA epochs plus archival post-upgrade VLA data: between 2016 March 8 (Epoch-1a) and 2016 April 7 (Epoch-1b), the Band 4 flux density dropped from about 2.2 mJy to about 0.9 mJy with comparable synthesized beams, then rose back to an intermediate 1.4–1.7 mJy level in the 2017 August epoch (Epoch-2). The roughly 30-day drop yields a rest-frame causality limit $R_{\\rm var} \\lesssim c\\,\\Delta t_{\\rm obs}/(1+z) \\simeq 0.024$ pc $\\simeq 1.1\\times10^3 R_{\\rm S}$ for the variable region, independent of any spectral decomposition. A joint spectral fit then requires a compact, curved, self-absorbed synchrotron component on top of extended star-formation and dust emission, and the hybrid thermal/nonthermal coronal model reproduces the two well-sampled epochs with different nonthermal electron indices ($p_{1a}=3.16$ versus $p_2=2.21$), while the intrahour light curves show only isolated excursions that the authors do not claim as secure variability. Radiative cooling times comparable to or shorter than the light-crossing time imply continuous injection or reacceleration of the electrons. The paper is explicit that the data do not uniquely distinguish a magnetized corona from a stratified corona, a corona-connected outflow, or a jet base.","pith_inferences":["If the causality bound is secure, single-band millimeter monitoring becomes a cheap, model-free probe of coronal scales in radio-quiet AGNs generally: any month-scale flux drop near 100 GHz brackets the emitting zone in Schwarzschild radii without spectral fitting, a survey strategy the paper demonstrates but does not itself propose.","The fitted shift in nonthermal index ($p \\approx 3.2$ to $\\approx 2.2$) resembles the spectral softening seen in accreting black-hole binaries near state transitions, but the authors' own error budget leaves the shift marginal; a discriminating test would be dense multi-band monitoring across a millimeter flare instead of isolated epoch pairs.","Because the data cannot separate a corona from a jet base or a stratified outflow, a natural extension is to search the growing ALMA sample of radio-quiet Seyferts for similar month-scale millimeter drops at the $\\sim10^3 R_{\\rm S}$ scale; if such events are common, the coronal interpretation gathers weight, whereas events that track jet scaling relations would not.","If the corona genuinely hosts the variable millimeter emitter, broadband snapshot campaigns of radio-quiet AGNs may frequently catch the millimeter and X-ray bands in mismatched states, since the hard X-ray flux did not track the millimeter changes here — a caution for interpreting single-epoch spectral energy distributions of these objects."],"forward_implications":["A radio-quiet AGN's compact millimeter emission can originate from a region of order $10^3$ Schwarzschild radii, placing it at the scale of the X-ray corona rather than the kiloparsec-scale extended emission.","The compact millimeter component is state-changing: its disappearance and reappearance at intermediate strength mean single-epoch detections of such excesses should not be treated as stable measures of an AGN's millimeter output.","Because the radiative cooling times of the millimeter-emitting electrons are comparable to or shorter than the light-crossing time, the nonthermal electron population must be continuously injected or reaccelerated, and a change in particle injection is the most natural driver of the observed spectral changes.","The lack of a clear correspondence between the millimeter states and the monthly Swift/BAT hard-X-ray light curve implies that population-level millimeter–X-ray correlations do not force one-to-one tracking on short timescales.","Filling the frequency gap between the VLA and ALMA bands would resolve the synchrotron turnover and test whether the compact emitter is a homogeneous one-zone source or has radial structure, a test the authors explicitly propose."],"supporting_citations":[{"why":"Supplies the Cycle 4 ALMA data used here as Epoch-2 and the hybrid-corona spectral decomposition framework that this paper updates and refits.","marker":"(Y. Inoue & A. Doi 2018)"},{"why":"Prior quasi-simultaneous radio observations establishing NGC 985's high-frequency excess and providing the black-hole mass used to scale the size limit.","marker":"(A. Doi & Y. Inoue 2016)"},{"why":"Introduces the hybrid thermal/nonthermal coronal synchrotron model used to interpret the compact millimeter component.","marker":"(Y. Inoue & A. Doi 2014)"},{"why":"Source of the black-hole mass adopted to express the causality size limit in Schwarzschild radii.","marker":"(M. Kim et al. 2008)"},{"why":"Provides the bolometric and X-ray context for NGC 985 used in the cooling-time and millimeter–X-ray ratio estimates.","marker":"(R. V. Vasudevan et al. 2009)"},{"why":"Supplies the particle acceleration–cooling balance picture behind the continuous-injection interpretation of the nonthermal electrons.","marker":"(Y. Inoue et al. 2019)"},{"why":"The millimeter–X-ray luminosity relation against which the epoch-specific compact millimeter fluxes of NGC 985 are compared.","marker":"(C. Ricci et al. 2023)"},{"why":"Day-scale millimeter variability limits in other radio-quiet AGNs, the comparison set for the causality-size argument.","marker":"(T. Michiyama et al. 2024)"}],"fun_headline_variants":["Month-scale drop pins NGC 985's millimeter excess to under 0.024 pc","Millimeter excess in NGC 985 varies on month timescales, hinting at corona","NGC 985's compact millimeter source dims and rebrightens with no spectral assumptions","ALMA shows recurrent compact millimeter synchrotron emission from NGC 985","NGC 985's variable millimeter compact component spans only 1,100 R_S"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing premise is that the roughly 1.3 mJy Band 4 drop between the two ALMA epochs one month apart is intrinsic source variability rather than a calibration artifact — a worry the paper itself acknowledges, since the absolute flux calibrator drifted from about 2.2 to 1.7 Jy over the same interval and the phase calibrator was sampled only two or three times per epoch.","fun_headline_variants_meta":{"raw":{"variants":["Month-scale drop pins NGC 985's millimeter excess to under 0.024 pc","Millimeter excess in NGC 985 varies on month timescales, hinting at corona","NGC 985's compact millimeter source dims and rebrightens with no spectral assumptions","ALMA shows recurrent compact millimeter synchrotron emission from NGC 985","NGC 985's variable millimeter compact component spans only 1,100 R_S"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000473,"raw_usage":{"cost":0.000473018,"is_byok":false,"cost_details":{"upstream_inference_cost":0.000473018,"upstream_inference_prompt_cost":3.4762e-05,"upstream_inference_completions_cost":0.000438256},"total_tokens":3662,"prompt_tokens":1254,"completion_tokens":2408,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":1090,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":2310}},"tokens_in":1254,"tokens_out":2408,"cost_source":"openrouter_usage","duration_ms":18404,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":2310,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T07:21:15.313969+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"Re-derive the Epoch-1a and Epoch-1b flux scales from an independent absolute calibrator or from ALMA's planet-based scale: if a $\\sim$1.3 mJy Band 4 difference between 2016 March 8 and April 7 survives, the intrinsic-variability claim and the $\\lesssim 0.024$ pc size limit stand; if it shrinks to the $\\sim$5% calibration floor, both collapse. A prospective check is weekly ALMA Band 4 monitoring of NGC 985 over several months: if the 150 GHz flux stays constant to within calibration noise, the month-scale drop was a singular event rather than a recurrent property, and the claim that NGC 985 hosts a recurrent variable component would need revision.","supporting_citations":[{"cited_title":"Detection of Coronal Magnetic Activity in nearby Active Supermassive Black Holes","cited_arxiv_id":null,"evidence_quote":"Supplies the Cycle 4 ALMA data used here as Epoch-2 and the hybrid-corona spectral decomposition framework that this paper updates and refits."},{"cited_title":", keywords =","cited_arxiv_id":null,"evidence_quote":"Prior quasi-simultaneous radio observations establishing NGC 985's high-frequency excess and providing the black-hole mass used to scale the size limit."},{"cited_title":"2014, Publications of the Astronomical Society of Japan, 66, L8, doi: 10.1093/pasj/psu079","cited_arxiv_id":null,"evidence_quote":"Introduces the hybrid thermal/nonthermal coronal synchrotron model used to interpret the compact millimeter component."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Source of the black-hole mass adopted to express the causality size limit in Schwarzschild radii."},{"cited_title":"On High-energy Particles in Accretion Disk Coronae of Supermassive Black Holes: Implications for MeV Gamma-rays and High-energy Neutrinos from AGN Cores","cited_arxiv_id":null,"evidence_quote":"Supplies the particle acceleration–cooling balance picture behind the continuous-injection interpretation of the nonthermal electrons."},{"cited_title":"A Tight Correlation between Millimeter and X-Ray Emission in Accreting Massive Black Holes from &lt;100 mas Resolution ALMA Observations","cited_arxiv_id":null,"evidence_quote":"The millimeter–X-ray luminosity relation against which the epoch-specific compact millimeter fluxes of NGC 985 are compared."},{"cited_title":"ALMA Confirmation of Millimeter Time Variability in the Gamma-Ray Detected Seyfert Galaxy GRS 1734-292","cited_arxiv_id":null,"evidence_quote":"Day-scale millimeter variability limits in other radio-quiet AGNs, the comparison set for the causality-size argument."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}