{"id":"6b6f84ba-330c-4528-892a-9c1bfa615617","arxiv_id":"2608.08369","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"JWST observations of five AGN reveal that radiation and shocks from the active nucleus produce opposite PAH signatures, explaining the bimodal PAH ratios in active galaxies.","lead":"JWST maps of five nearby active galaxies show two distinct ways the central black hole affects surrounding dust and gas, leaving opposite signatures in PAH emission. The findings give astronomers a new tool to identify AGN feedback mode and improve star formation rate measurements.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quantitative evidence that irradiation and shocks collectively suppress PAH emission is weakened by mathematical coupling: the predictor ratios in Eq. (1) contain the same PAH fluxes being fit.","rationale":"The paper's central claim has two pillars: the regression relating PAH suppression to SFR, rH2, and rNe (Section 3.1, Eq. 1), and the bimodal trends in PAH ratio diagrams (Figures 4 and 5). The regression pillar is the only quantitative test that simultaneously attributes PAH suppression to both feedback modes, and it is compromised by mathematical coupling: the predictor ratios contain the very PAH features being predicted. This is a concrete, checkable flaw, and the proposed refit with absolute tracers would settle whether the fitted coefficients are physical or algebraic. I do not think this alone overturns the qualitative framework, because Figure 5 uses ionized-line ratios that do not share PAH bands and because the spatial associations (radio spots, ionization cones) provide partially independent evidence. However, the paper's language in the abstract overstates the quantitative support. The reader's weakest_assumption focused on rH2 and rNe not being clean tracers; that is a distinct but related concern. The coupling issue is more fundamental because it affects the regression even if the tracers were clean. The target selection based on nuclear PAH properties spanning two regimes (Section 2.1) also weakens the bimodality as independent evidence, but the within-target spatial gradients and literature comparison carry some weight. On balance, the conditional verdict remains appropriate: the framework is plausible and well-illustrated, but the quantitative demonstration needs repair before the method is adopted as a calibrated diagnostic.","tokens_in":31184,"tokens_out":6998,"duration_ms":74329,"concrete_test":"Refit Eq. (1) replacing rH2 and rNe with absolute surface-brightness predictors log H2 S(1–5) and log [NeV] (keeping log Σ_SFR as a control), and repeat the Figure 4 color-coding using those absolute tracers normalized by Σ_SFR rather than by a PAH band. If α2 and α3 remain significantly negative and the two color-coded trends persist, the coupling explanation fails and the claim is supported. If the coefficients flip sign or lose significance, the current regression evidence is an artifact of shared denominators.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing assumption in the quantitative core is that the two predictor ratios in Eq. (1) can be treated as independent physical tracers even though they share the dependent PAH features. For PAH 7.7, the model fits log Σ_PAH,7.7 against rH2 = H2 S(1–5)/PAH7.7; for PAH 11.3 it fits against rNe = [NeV]/PAH11.3. Any spaxel with a low PAH flux—whether because of AGN irradiation, shocks, extinction mismatch, or noise—automatically has a high rH2 (or rNe) at fixed H2 (or [NeV]). The large negative coefficients in Table A1 (α2 ≈ −0.86 to −0.90 for rH2; α3 ≈ −0.19 to −0.32 for rNe) are therefore partly algebraic, not physical. Figures 4 and 5 inherit the same issue where the color-coding variables are built from PAH7.7 or PAH11.3, the same bands appearing on the axes, so the apparent two trends can be manufactured by shared denominators. The paper's admitted complication in Circinus—where rH2 is consistent with a significant star-formation contribution to H2 excitation (Section 4.2)—adds an independent confound. Until the coupling is removed, Eq. (1) and the associated RMS scatters do not establish that the two modes collectively drive PAH suppression. The [FeII]/[ArII] and [NeV]/[NeIII] proxies in Figure 5 are not coupled to PAH and provide partially independent support, but the quantitative regression and the main color-coding in Figure 4 remain compromised.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents JWST MIRI/MRS spatially resolved spectroscopy of the central regions of five AGN (three LINERs and two Seyferts) and analyzes PAH, H2, and neon/iron line diagnostics at ~4--24 pc scales. The central claim is that two AGN feedback modes, radiative (AGN irradiation) and kinetic (shock processing), jointly suppress PAH emission and that their relative importance explains the bimodal distribution of PAH band ratios in these galaxies and reconciles apparently conflicting results in the literature. The quantitative support is based on: (i) a multiple linear regression (Eq. 1) that reconstructs PAH surface brightness from SFR, rH2 = H2 S(1--5)/PAH7.7, and rNe = [NeV]/PAH11.3; (ii) spatially resolved maps and PAH ratio diagrams, with model grids for neutral and ionized PAHs; and (iii) independent ionized-line diagnostics, particularly [FeII]/[ArII] and [NeV]/[NeIII]. The paper also proposes a practical framework for calibrating SFRs from PAH emission in AGN and for diagnosing AGN feedback in JWST-era observations.","tokens_in":31707,"tokens_out":6860,"duration_ms":65849,"significance":"If the central claim holds, the paper provides a genuinely useful observational framework: PAH band ratios, when combined with shock and irradiation tracers, could encode the dominant feedback mode in AGN, with implications for SFR calibration and for subgrid prescriptions in galaxy-formation simulations. The manuscript has notable strengths: the reduction is careful, with PSF matching, spaxel-based MCMC decomposition, and aperture robustness checks; the sample spans a wide range in L_bol; and the independent ionized-line diagnostics in Figure 5 and Appendix A2 provide partially non-circular support for the qualitative picture. These strengths are, however, tempered by a load-bearing statistical problem: the main regression and the main color-coding variables are built from ratios that contain the same PAH fluxes that are being explained, so the tight correlations in Figure 2 and the apparent trends in Figure 4 are partly algebraic rather than physical.","major_comments":[{"comment":"The multiple linear regression is subject to direct mathematical coupling. For PAH7.7, the dependent variable is log Σ_PAH,7.7 and one predictor is log rH2 = log H2 − log PAH7.7; rearranging Eq. (1) puts (1+α2) log PAH7.7 on the left, and with α2 ≈ −0.86 this leaves only a residual coefficient of about 0.14. The regression can therefore produce a tight fit and a large negative α2 even if H2 and PAH7.7 are only weakly related physically. Similarly, for PAH11.3 the predictor rNe contains PAH11.3 in the denominator while PAH11.3 is on the left-hand side, so α3 ≈ −0.32 is partially built in. The RMS scatters in Figure 2 are consequently not a valid measure of how well the physical model reproduces PAH suppression. The same issue affects Eq. (2) and its stated predictive scatter. Please re-run the analysis with predictors that do not contain the dependent flux, for example SFR, H2 S(1--5), and [NeV] as separate terms, or H2/[NeII] and [NeV]/[NeII]; and provide a null test (e.g., shuffled PAH fluxes or simulated noise with the same covariance) to demonstrate that the improved fit is not reproduced by pure coupling.","section":"§3.1, Eq. (1), Table A1"},{"comment":"The main color-coding in Figure 4 reproduces the concern. In panel (a) the color variable rH2 = H2/PAH7.7 shares the PAH7.7 denominator with both the x-axis (PAH6.2/PAH7.7) and the y-axis (PAH11.3/PAH7.7), so low-PAH7.7 spaxels simultaneously move to the upper right and receive high rH2 values; this can manufacture the apparent shock-processing sequence. In panel (b), rNe = [NeV]/PAH11.3 shares PAH11.3 with the y-axis numerator, so a spaxel with weak PAH11.3 appears both as a low PAH11.3/7.7 value and as a high-irradiation point. The independent proxies in Figure 5 do not share denominators with the PAH ratio axes and partially mitigate this problem, but the main quantitative demonstration in Figure 2 and the primary visual separation in Figure 4 remain compromised. Please make the independent-proxy version the primary figure, or demonstrate with a null simulation that the binned trends in Figure 4 are not produced by shared denominators.","section":"Figure 4"},{"comment":"The paper states that the rH2 values of most spaxels in Circinus are consistent with a significant contribution from star-formation activity to the excitation of H2 (Section 4.2). This directly undercuts the assumption, used throughout Section 3, that rH2 is a clean tracer of shock processing. Since Circinus is one of only two Seyferts in the sample and contributes to the irradiation-dominated arm, this confound affects the separation of the two modes in Figures 3 and 4. Please quantify the sensitivity of the results to this issue: exclude Circinus from the regression and from the median-bin trends, or model the star-formation contribution to H2 excitation, and show that the shock-processing arm is not an artifact of SF-heated H2 in (post-)starburst systems.","section":"§4.2 (Circinus); §3.1"},{"comment":"The statistical analysis treats individual spaxels as independent data points, but the five galaxies occupy different regions of the parameter space and neighboring spaxels are spatially correlated. The quoted coefficient uncertainties and RMS scatters therefore overstate the effective sample size. In addition, the claim of a bimodal distribution of PAH 11.3/7.7 ratios is made on the basis of visual separation around a demarcation value of 0.3, without a formal two-component test. Please report leave-one-galaxy-out cross-validation or cluster-bootstrap errors for the regression coefficients, and apply a standard bimodality test (e.g., Hartigan's dip test) to the PAH ratio distribution before describing it as bimodal.","section":"§3.1, Table A1; §3.2"}],"minor_comments":[{"comment":"There is a typo in the opening sentence: “ionzied gas phases” should be “ionized gas phases.”","section":"§3.1"},{"comment":"In the introduction, “UV-optical-infared emission lines” should read “UV-optical-infrared emission lines.”","section":"§1"},{"comment":"The phrase “exhibit the the same bimodal distribution” contains a duplicated article; please correct it.","section":"§3.2"},{"comment":"The axis label “log (H2 S(1 5)/PAH7.7)” is missing the en dash in S(1--5); please fix the formatting for consistency with the text.","section":"Figure A1 caption"},{"comment":"The description of the pipeline options (“master bg function”, “firstframe option”) would be clearer if the actual calibration pipeline parameter names were given in a monospaced style, since these are implementation details used by other JWST users.","section":"§2.2"}],"recommendation":"major_revision","confidential_remarks":"The circularity in Eq. (1) and Figure 4 is, in my view, the central obstacle. The authors may be able to fix it within the manuscript's scope by replacing the coupled predictors with independent ones, adding a null test, and re-running the two key figures with the ionized-line proxies from Figure 5. The qualitative picture is plausible and the data are valuable, but as written the quantitative evidence for the headline claim is not yet established. I would also ask the authors, in revision, to clarify the incremental novelty relative to previous GATOS papers (especially Zhang et al. 2026), since the present analysis builds directly on that work."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's my read: the paper is worth reading for the data alone, and the framework is plausible. The quantitative regression evidence, though, is weakened by mathematical coupling.\n\nWhat's actually new: first spatially resolved MIRI/MRS multiphase analysis of these five AGN at 4–24 pc scales, including new observations of NGC 7314 and Circinus. They combine PAH, H2, and neon diagnostics and argue that radiative (irradiation) and kinetic (shock) feedback suppress PAH emission in distinct ways, producing a bimodal distribution of PAH band ratios. The reconciliation of prior conflicting results on PAH-based SFR calibrators is genuinely interesting and testable.\n\nThe paper does a lot well. Reduction is careful: PSF matching, Bayesian PAH decomposition with tau_9.7, perturbation-based line fitting. They are candid about known issues — Circinus saturation, possible star-formation contribution to H2 excitation, factor-of-two method-dependent PAH flux differences. The qualitative trends in Figures 4 and 5 survive when color-coded by independent line ratios ([NeV]/[NeIII], [FeII]/[ArII]) that do not involve PAH in the denominator, so the core picture does not collapse.\n\nThe main soft spot is the quantitative core. Equation 1 regresses log(PAH7.7) against log(rH2) = log(H2/PAH7.7), and log(PAH11.3) against log(rNe) = log([NeV]/PAH11.3). That's mathematical coupling: low PAH flux at fixed H2 or [NeV] automatically inflates the predictor. The large negative coefficients in Table A1 are therefore partly algebraic, and Figure 4's color-coding inherits the same shared denominator. The independent proxies in Figure 5 mitigate this, but the tight scatter in Figure 2 should not be read as independent confirmation of the \"collectively drive\" claim. Also, targets were selected on the basis of their known PAH properties, so the bimodality is not an unbiased discovery.\n\nIn short: framework is plausible, data are valuable, and the qualitative case is fairly strong despite the regression flaw. I would send it to peer review, but I would require the authors to refit Equation 1 using predictors that do not share the dependent PAH flux (e.g., H2/SFR, [NeV]/SFR, or the line-ratio proxies from Figure 5), and to discuss the sample selection explicitly as a limitation rather than as support. A serious referee will get a useful paper out of this.","headline":"Genuinely new JWST data and a plausible unified framework for PAH suppression in AGN, but the headline regression is partly circular and the qualitative case rests on the independent line-ratio probes.","tokens_in":32265,"tokens_out":2748,"would_cite":true,"duration_ms":27613,"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":"JWST spectroscopy of five AGN shows radiative and kinetic feedback leave opposite, identifiable signatures in PAH molecules, and their coexistence explains both the observed bimodal PAH ratios and earlier conflicting results.","keywords":["AGN feedback","PAH emission","shock processing","AGN irradiation","JWST MIRI/MRS spectroscopy","molecular hydrogen","bimodal PAH band ratios","star formation rate calibration"],"falsifier":"A targeted test: in a sample of AGN with resolved radio jets and known ionization cones, the framework predicts 11.3/7.7 > 0.3 along the jet axis and < 0.3 inside the cone for the same galaxy; a single galaxy showing the opposite spatial association would falsify the scheme.","tokens_in":31026,"feed_emoji":"🔭","tokens_out":11678,"duration_ms":89009,"temperature":0.7,"pith_summary":"Using JWST MIRI/MRS spectra of the central ~40–240 pc of five active galactic nuclei, this paper argues that the two classic AGN feedback modes—radiative (irradiation) and kinetic (shock processing)—leave opposite, identifiable fingerprints on polycyclic aromatic hydrocarbon (PAH) molecules. In the three LINERs, spaxels with strong shock indicators show elevated PAH 11.3/7.7 ratios, the signature of large, neutral PAHs; in the two Seyferts, strong irradiation indicators accompany low 11.3/7.7 and 6.2/7.7 ratios, the signature of large, ionized PAHs. The two modes coexist in the nuclear regions, with one generally dominating. The paper claims this coexistence explains both the bimodal PAH band-ratio distribution it maps and the conflicting literature on which PAH feature is a reliable star-formation tracer. If right, the physical state of PAHs becomes a practical diagnostic of which AGN feedback mode is at work, with direct consequences for calibrating star-formation rates in AGN hosts.","feed_headline":"Two AGN feedback modes leave opposite imprints on PAH molecules","feed_subtitle":"PAH size and charge encode which feedback dominates, reconciling disputed SFR calibrations","key_machinery":"The load-bearing tools are two spaxel-based diagnostic ratios—$r_{\\rm H_2}={\\rm H_2\\,S(1\\!-\\!5)}/{\\rm PAH\\,7.7}$, adopted as a shock-processing indicator because shock heating creates excess H$_2$ emission relative to PAHs, and $r_{\\rm Ne}=[{\\rm Ne\\,V}]14.32\\,\\mu{\\rm m}/{\\rm PAH\\,11.3}$, adopted as an AGN-irradiation indicator because the coronal $[{\\rm Ne\\,V}]$ line tracks AGN luminosity—combined with the multiple linear regression $\\log\\Sigma_{\\rm PAH}=\\alpha_1\\log\\Sigma_{\\rm SFR}+\\alpha_2\\log r_{\\rm H_2}+\\alpha_3\\log r_{\\rm Ne}+\\beta$. These ratios are read against the PAH 6.2/7.7 versus 11.3/7.7 band-ratio diagram, where theoretical grids (neutral, 70% ionized, and fully ionized PAHs with carbon numbers $N_C=20$–$400$ at interstellar radiation fields up to $10^3$ times the local value) convert observed ratios into PAH size and ionization state. The explanatory mechanism is differential destruction: shocks selectively destroy small and ionized PAHs, while irradiation ionizes PAHs and preferentially destroys small grains.","core_discovery":"The central claim is that PAH band ratios—specifically PAH 6.2/7.7 versus 11.3/7.7—separate cleanly into two tracks across the spaxels of five AGN. One track, driven by shock processing (traced by ${\\rm H_2\\,S(1\\!-\\!5)}/{\\rm PAH\\,7.7}$ and $[{\\rm Fe\\,II}]5.34\\,\\mu{\\rm m}/[{\\rm Ar\\,II}]6.99\\,\\mu{\\rm m}$), runs toward higher 11.3/7.7 and reflects a population of larger, neutral PAHs, because shocks preferentially destroy small and ionized PAHs. The other track, driven by AGN irradiation (traced by $[{\\rm Ne\\,V}]14.32\\,\\mu{\\rm m}/{\\rm PAH\\,11.3}$ and $[{\\rm Ne\\,V}]/[{\\rm Ne\\,III}]$), runs toward lower 11.3/7.7 and 6.2/7.7 and reflects larger, more ionized PAHs, because extreme-UV and X-ray photons ionize and preferentially destroy smaller grains. A multiple linear regression including SFR surface density and both ratios reproduces the observed PAH surface brightness with 0.19–0.25 dex scatter, showing the suppression is jointly governed by the two modes. The paper further argues that the bimodal 11.3/7.7 distribution (demarcated at 0.3) and the apparently contradictory findings of earlier work are reconciled once sample selection is accounted for: shock-dominated nuclei behave like one earlier sample, irradiation-dominated nuclei like another.","pith_inferences":["A magnitude-limited survey of AGN spaxels classified by $r_{\\rm H_2}$ and $r_{\\rm Ne}$ should show a visibly double-peaked distribution in 11.3/7.7; this can be tested with existing archival JWST data without new observations.","Because PAH destruction operates on roughly $10^3$–$10^5$ yr timescales (as the paper notes), combining PAH-state maps with instantaneous coronal-line strength could measure AGN duty cycles, an application the paper raises but does not develop.","The $r_{\\rm H_2}$ shock indicator is vulnerable to star-formation contamination, as the Circinus case shows; a natural extension is to calibrate it spaxel-by-spaxel against the H$_2$ temperature-distribution slope $\\beta$ and exclude spaxels where $\\beta$ indicates photoelectric heating rather than shocks.","If the framework holds, corrected PAH 7.7 fluxes would give SFRs in AGN accurate to roughly 0.2 dex without optical Balmer extinction corrections, simplifying SFR estimates in large AGN surveys."],"forward_implications":["PAH 7.7 µm emission corrected with the two-ratio formula becomes the tightest-correlated PAH-based SFR tracer in AGN, with 0.19 dex scatter in this sample.","Relative PAH suppression at a given SFR can be decomposed into an irradiation term and a shock term, so a PAH deficit no longer has an ambiguous physical origin.","Nuclei dominated by shocks (high 11.3/7.7, high $r_{\\rm H_2}$) and nuclei dominated by irradiation (low ratios, high $r_{\\rm Ne}$) follow opposite PAH–SFR behaviors, explaining why prior studies concluded different PAH features were the reliable SFR tracer.","Shock processing from jets or outflows is not confined to LINERs: all NGC 7314 spaxels with 11.3/7.7 > 0.4 lie above the shock-contribution threshold in $r_{\\rm H_2}$, showing Seyferts also carry kinetic-mode imprints.","The framework lays the groundwork for a quantitative AGN feedback diagnostic based on PAH features, with future anchor points and a mixing-sequence PAH diagram planned from larger samples."],"supporting_citations":[{"why":"Prior JWST analysis of the three LINERs establishing the H2 excitation, slow C-type shock velocities, and the high 11.3/7.7 shock-processing interpretation that this paper extends to a larger sample.","marker":"Zhang et al. 2026"},{"why":"Supplies the theoretical PAH model grids (neutral, 70% ionized, fully ionized, NC=20–400) used to translate observed PAH band ratios into PAH size and ionization state.","marker":"Rigopoulou et al. 2024"},{"why":"Provides the unity PAH–SFR relations used as the star-forming baseline against which AGN PAH suppression is measured.","marker":"Shipley et al. 2016"},{"why":"Provides the AGN-calibrated neon-line prescription used to derive spaxel SFR surface densities independent of PAH emission.","marker":"Zhuang et al. 2019"},{"why":"Supplies the theoretical models of H2 excitation used to identify C-type molecular shocks with vs ≲ 10 km/s and to interpret the H2 line ratios.","marker":"Kristensen et al. 2023"},{"why":"Theoretical destruction rates of PAHs by shocks underpin the interpretation that shocks preferentially destroy small and ionized PAHs.","marker":"Micelotta et al. 2010a"},{"why":"JWST spectroscopy of Seyfert nuclei showing elevated PAH 11.3/7.7 and H2/PAH ratios along outflows and radio jet axes, supporting shock processing in Seyferts.","marker":"García-Bernete et al. 2024b"},{"why":"JWST measurements of Seyfert apertures with PAH 11.3/7.7 > 0.4 and log rH2 values comparable to the LINERs, providing literature evidence for the unified framework.","marker":"Zhang et al. 2024b"},{"why":"One of the conflicting results the framework reconciles, concluding that PAH 11.3 remains a robust SFR tracer in shock-dominated nuclei.","marker":"Diamond-Stanic & Rieke 2010"},{"why":"The other conflicting result the framework reconciles, finding PAH 11.3 suppressed relative to 7.7 in irradiation-dominated AGN.","marker":"LaMassa et al. 2012"}],"fun_headline_variants":["JWST spies twin AGN feedback modes in PAH emission","Two AGN feedback tracks explain PAH ratio bimodality","Shocks and radiation leave distinct PAH fingerprints in AGN","PAH ratios expose radiative vs kinetic AGN feedback","JWST maps AGN feedback: PAH ratios split by mode"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The case rests on the premise that the H2-to-PAH ratio mainly measures shock heating and the [NeV]-to-PAH ratio mainly measures AGN irradiation; the paper itself notes Circinus's H2 may be substantially excited by star formation, so if that contamination is widespread the two tracks and the bimodal interpretation weaken.","fun_headline_variants_meta":{"raw":{"variants":["JWST spies twin AGN feedback modes in PAH emission","Two AGN feedback tracks explain PAH ratio bimodality","Shocks and radiation leave distinct PAH fingerprints in AGN","PAH ratios expose radiative vs kinetic AGN feedback","JWST maps AGN feedback: PAH ratios split by mode"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000651,"raw_usage":{"total_tokens":3097,"prompt_tokens":1171,"completion_tokens":1926,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":787,"completion_tokens_details":{"reasoning_tokens":1841}},"tokens_in":787,"tokens_out":1926,"duration_ms":12576,"temperature":1.0,"reasoning_tokens":1841,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:05:52.349939+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A targeted test: in a sample of AGN with resolved radio jets and known ionization cones, the framework predicts 11.3/7.7 > 0.3 along the jet axis and < 0.3 inside the cone for the same galaxy; a single galaxy showing the opposite spatial association would falsify the scheme.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior JWST analysis of the three LINERs establishing the H2 excitation, slow C-type shock velocities, and the high 11.3/7.7 shock-processing interpretation that this paper extends to a larger sample."},{"cited_title":"R., Garc´ıa-Bernete, I., et al","cited_arxiv_id":null,"evidence_quote":"Supplies the theoretical PAH model grids (neutral, 70% ionized, fully ionized, NC=20–400) used to translate observed PAH band ratios into PAH size and ionization state."},{"cited_title":"E., Godard, B., Guillard, P., et al","cited_arxiv_id":null,"evidence_quote":"Supplies the theoretical models of H2 excitation used to identify C-type molecular shocks with vs ≲ 10 km/s and to interpret the H2 line ratios."},{"cited_title":"M., Heckman, T","cited_arxiv_id":null,"evidence_quote":"The other conflicting result the framework reconciles, finding PAH 11.3 suppressed relative to 7.7 in irradiation-dominated AGN."}],"review_version":1}