{"id":"561bc17c-7628-4c70-a575-5f1c6ca638d3","arxiv_id":"2607.05249","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":8,"one_line_summary":"In 24 BHXRB outbursts, the thermal disk luminosity drops below the soft-state exponential decay baseline, and this deficit is used as a tracer of disk truncation, corroborated by decreasing timing frequencies.","lead":"This paper uses the drop in thermal disk X-ray luminosity below an exponential decay baseline as a tracer for the thin accretion disk receding during soft-to-hard state transitions in 24 black hole X-ray binary outbursts. A smart generalist might read it to understand how astronomers infer changes in accretion geometry when direct measurements fail, and how timing signals cross-check the result.","discovery_kind":"unclear","skeptic_critique":{"model":"glm-5.2","headline":"The tracer's validity hinges on unverified exponential mass inflow persistence; a total-luminosity check against the baseline would directly test this.","rationale":"The reader identified the correct load-bearing concern: the assumption that Ṁ continues exponentially after the break is unverified and central to the tracer's validity. I agree this is the weakest link. The paper is appropriately transparent — it frames the result as a 'tracer' rather than a measurement, discusses alternatives in Section 5.2, and acknowledges that coronal dissipation changes or outflows could also produce a luminosity deficit without geometrical recession. The timing cross-check provides qualitative support but does not independently verify the Ṁ assumption, since both the recession tracer and timing frequencies respond to the same transition. The circularity the reader notes (Eq. 4 defines the tracer by construction) is real but is mitigated by the paper's honest framing. The concrete test I propose — checking whether L_total = L_disk + L_Comp tracks the exponential baseline — is the most direct way to validate or falsify the mass inflow assumption using data the authors already have. It exploits the physical prediction that if Ṁ persists exponentially and the disk recedes, the accretion power should be redistributed to the Comptonized component rather than lost. If this test passes, the CONDITIONAL verdict could be upgraded; if it fails, the tracer's physical interpretation is undermined. The verdict remains CONDITIONAL because the concern is genuine but the paper does not overclaim, and the proposed test could resolve the issue without new observations.","tokens_in":24233,"tokens_out":1946,"duration_ms":90391,"concrete_test":"For each outburst in the sample, compute L_total(t) = L_disk(t) + L_Comp(t) using the flux decomposition in Section 2.2, and compare L_total to the extrapolated exponential baseline L_exp(t) after t_end. If L_total/L_exp remains consistent with unity (within the propagated uncertainties), the mass inflow assumption is supported and the deficit genuinely traces redistribution of accretion power (consistent with disk recession). If L_total also drops significantly below L_exp, the deficit reflects a decline in mass supply rather than geometry, and Eq. 4 is not a valid recession tracer. This test requires no new data — only combining the already-measured disk and Comptonized fluxes shown in Figures 8-9.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader correctly identifies the load-bearing assumption: Eq. 4 (R_tr/R_ISCO = L_exp/L_obs) is only valid if the mass inflow rate Ṁ continues to follow the soft-state exponential decay after t_break (Section 3). If Ṁ itself deviates — due to outer-disk cooling, irradiation loss, or changing mass supply — the luminosity deficit conflates geometry with mass supply, and the inferred R_tr/R_ISCO is not a tracer of disk recession but of whatever causes Ṁ to drop. The paper acknowledges this in Section 5.2 and performs an order-of-magnitude irradiation radius check (Figure 4), arguing that irradiation should still sustain the outer disk at the onset of the deficit. However, this check only addresses one specific alternative (complete irradiation loss); it does not verify that Ṁ actually continues exponentially. The timing cross-check (Section 4.2) is qualitative and does not independently constrain Ṁ, since both the recession tracer and the timing frequencies are driven by the same underlying transition. The concern is real but the paper is transparent about it, which is why CONDITIONAL is appropriate rather than REJECT. The key untested prediction is: if Ṁ continues exponentially while the disk recedes, the accretion power should be redistributed from the thermal disk to the Comptonized component, so the total luminosity L_disk + L_Comp should approximately track the extrapolated exponential baseline. The paper shows the Comptonized component rising as the disk drops (Figures 8-9) but never explicitly checks whether the sum tracks L_exp(t).","agreement_with_reader":"agree"},"referee_report":{"model":"glm-5.2","summary":"The manuscript presents a systematic study of 26 BHXRB outbursts observed with RXTE/PCA, proposing that the thermal disk luminosity deficit relative to an extrapolated soft-state exponential baseline serves as an empirical tracer of thin disk recession during soft-to-hard transitions. The central relation (Eq. 4) defines R_tr/R_ISCO = L_exp/L_obs, where L_exp is the extrapolated exponential and L_obs is the observed disk luminosity. The authors apply this tracer to 24 outbursts, finding that the inferred truncation radius increases during the transition, and present a qualitative timing cross-check showing decreasing characteristic frequencies. The paper also discusses the diversity of recession behavior across sources and addresses the distinction between irradiation-driven decay and disk evaporation.","tokens_in":24424,"tokens_out":1731,"duration_ms":379545,"significance":"The paper addresses a genuine observational challenge — the unreliability of direct spectral radius measurements during soft-to-hard transitions — and proposes a practical, broadly applicable tracer using archival RXTE data across a large sample. The systematic nature of the analysis (24 outbursts, uniform spectral decomposition) and the MCMC-based baseline fitting procedure (Appendix B) are commendable. The qualitative timing cross-check and the irradiation radius consistency check (Section 5.2, Figure 4) add supporting evidence. The framework is transparent about its assumptions and limitations, which is appropriate. The falsifiable prediction — that total luminosity (disk + Comptonized) should approximately track the exponential baseline if mass inflow persists — is implicitly present in the data (Figures 8–9 show the Comptonized component rising as the disk drops) but is not explicitly tested as a quantitative check.","major_comments":[{"comment":"Section 3, Eq. (4): The recession tracer R_tr/R_ISCO = L_exp/L_obs is valid only if the mass inflow rate continues to follow the soft-state exponential decay after t_break. The paper acknowledges this assumption (Section 3, following B. You et al. 2023) and discusses it in Section 5.2, but the tracer's central claim depends on it. A direct, quantitative test is available but not performed: if the mass inflow persists exponentially while the disk recedes, the total luminosity L_disk + L_Comp should approximately track the extrapolated baseline (since accretion power is redistributed from the thermal disk to the Comptonized component). The individual light curves in Appendix C (Figures 8–9) show the Comptonized component rising as the disk drops, but no systematic comparison of L_total vs. L_exp is presented across the sample. Adding such a check — even for a representative subset — would ","section":null},{"comment":"Section 4.2, Figure 2: The timing cross-check is described as qualitative, and the authors note that frequencies 'should not be interpreted as direct measurements of the local Keplerian frequency.' However, the timing analysis is presented as a key corroborating result (point 2 of the Conclusion). The scatter in Figure 2 is substantial (spanning roughly an order of magnitude in frequency at any given R_tr/R_ISCO), and no quantitative correlation coefficient or fit is reported. A Spearman rank correlation or similar quantitative measure for the frequency–radius relation would strengthen the claim that the timing evolution is 'consistent with' disk recession, or alternatively would clarify the limits of the agreement.","section":null},{"comment":"Section 5.1, Eq. (6) and Figure 3B: The recession slope gamma and the break scale m_dot_break are both derived from the same exponential baseline, and the authors note they are 'not statistically independent quantities.' The reported Spearman r_S = 0.3 at 0.6 sigma significance is therefore difficult to interpret. While the authors frame this as exploratory, the conclusion states that gamma 'shows no statistically significant correlation with m_dot_break' as a substantive finding (point 3). Given the shared dependence on the fitted baseline, this non-correlation may reflect the statistical structure of the construction rather than a physical result. The authors should clarify whether the non-independence could artificially suppress or inflate the correlation, or restrict the claim to a more cautious statement.","section":null}],"minor_comments":[{"comment":"Abstract: '26 BHXRBs' vs. '24 outbursts' — the abstract states 26 sources were studied and 24 outbursts identified with the deficit. Table 1 lists 26 sources but Table 2 lists 32 outburst episodes fitted. The relationship between '26 BHXRBs,' '24 outbursts,' and the 32 entries in Table 2 should be clarified.","section":null},{"comment":"Section 2.2: The electron temperature is fixed at kT_e = 150 keV. Appendix Figure 5 shows the comparison for 4U 1543-47, but it would be useful to state how many sources were checked and whether any showed significant differences, rather than only one example.","section":null},{"comment":"Section 2.2: The diskbb normalization depends on the color correction factor f_col and the inclination, neither of which is discussed in the context of the luminosity-to-radius conversion. Since Eq. (4) uses luminosity ratios for the same source, f_col cancels, but the disk fluxes themselves depend on the spectral model assumptions. A brief note on this cancellation would help the reader.","section":null},{"comment":"Figure 1: The legend lists 24 outbursts but the panel B y-axis starts at 10^0. Some points are plotted as lower limits at R_tr = R_ISCO; the criteria for when a point is a lower limit vs. a measurement should be stated more precisely (currently described only as 'when the observed disk luminosity is higher than, or statistically consistent with, the extrapolated baseline').","section":null},{"comment":"Table 2: GS 1354-64 (1997) has tau = 48.4 +309.1/-5.0 days, which is extremely poorly constrained. It would be useful to flag such cases or note whether they are included in the Figure 1/3 analysis.","section":null},{"comment":"Section 5.2: The irradiation radius check (Figure 4) uses C_irr = 5e-3. The sensitivity of R_irr to this choice is not discussed; a brief note on how much R_irr changes for plausible C_irr values (e.g., 1e-3 to 1e-2) would strengthen the order-of-magnitude argument.","section":null},{"comment":"References: Several 2026 references (König et al. 2026; Zdziarski et al. 2026a,b; You et al. 2026) are cited. Ensure these are properly published or have stable arXiv identifiers at the time of submission.","section":null},{"comment":"Figure 2: The theory curves for Lense-Thirring precession are shown for a = 0.1, 0.5, 0.9 at 10 M_sun. The assumed hot-flow geometry (radial extent, surface density profile) is mentioned as important but not specified. A brief statement of the assumed geometry would make the comparison more interpretable.","section":null}],"recommendation":"major_revision","confidential_remarks":"The core idea is reasonable and the systematic sample is a strength, but the paper currently sits between a methods paper and a results paper without fully committing to either. The total-luminosity check (major comment 1) is the most important addition: it is a straightforward test that directly addresses the load-bearing assumption, and the data to perform it are already in hand. Without it, the tracer remains a definition rather than a validated tool. The timing cross-check (major comment 2) is too qualitative to carry the weight assigned to it in the conclusion. I would encourage the authors to either strengthen the timing analysis quantitatively or reframe it more explicitly as a consistency check rather than corroboration. The paper is appropriate in scope for the journal once these issues are addressed."},"author_rebuttal":{"model":"glm-5.2","summary":"We thank the referee for a careful and constructive report. All three major comments are well-taken and will be addressed in the revised manuscript. We agree to add a quantitative total-luminosity consistency check, report a formal correlation coefficient for the timing–radius relation, and soften the conclusion regarding the gamma–m_dot_break non-correlation. One standing limitation remains: the non-independence of gamma and m_dot_break cannot be fully disentangled without dedicated simulations, and we will state this explicitly.","responses":[{"response":"We agree that this is the most direct internal consistency check available for our central assumption, and we appreciate the referee identifying it. We will add a quantitative comparison of L_total = L_disk + L_Comp against the extrapolated exponential baseline L_exp for the full sample (or at minimum a representative subset), presented as a new figure and accompanying statistics in the revised manuscript. We note two caveats that we will also state explicitly in the paper. First, exact tracking of L_exp by L_total is a necessary but not sufficient condition: if the radiative efficiency of the hot inner flow differs from that of the thin disk at the ISCO, or if a fraction of the accretion power is carried away by outflows, L_total will deviate from L_exp even if the mass inflow rate continues exponentially. Second, the Comptonized flux in our spectral decomposition depends on the assumed electron temperature (fixed at kT_e = 150 keV) and on the thcomp covering fraction, introducing additional model-dependent uncertainty into L_total. Nevertheless, the test provides a valuable order-of-magnitude check: if L_total drops well below L_exp across the sample, the assumption of persistent exponential inflow would be seriously challenged. If L_total remains broadly consistent with L_exp (within the expected efficiency-factor scatter), the assumption is supported. We will report the results honestly regardless of outcome.","revision_made":"yes","referee_comment":"Section 3, Eq. (4): The recession tracer R_tr/R_ISCO = L_exp/L_obs is valid only if the mass inflow rate continues to follow the soft-state exponential decay after t_break. A direct, quantitative test is available but not performed: if mass inflow persists exponentially while the disk recedes, L_disk + L_Comp should approximately track the extrapolated baseline. No systematic comparison of L_total vs. L_exp is presented across the sample."},{"response":"This is a fair point. We will compute and report the Spearman rank correlation coefficient (with Monte Carlo propagation of measurement uncertainties, following the same procedure already used for the gamma–m_dot_break comparison) for the characteristic frequency versus R_tr/R_ISCO relation, separately for the broadband noise and QPO components shown in Figure 2. We expect the coefficient to confirm a negative trend but with substantial scatter, and we will report the significance level transparently. We will also revise the language in the Conclusion to accurately reflect the strength of the correlation rather than describing it only as 'qualitatively consistent.' If the correlation turns out to be weak in a statistical sense, we will say so and discuss the possible reasons (e.g., that the characteristic frequency is not a direct measurement of the Keplerian frequency at the truncation radius, that different variability components may trace different radii, and that red-noise leakage and multi-component decomposition introduce additional scatter). The timing analysis will remain a supporting consistency check rather than a primary result, consistent with how it is framed in the manuscript.","revision_made":"yes","referee_comment":"Section 4.2, Figure 2: The timing cross-check is qualitative, but presented as a key corroborating result (Conclusion point 2). The scatter is substantial and no quantitative correlation coefficient is reported. A Spearman rank correlation would strengthen or clarify the claim."},{"response":"The referee raises a valid concern. We already note in Section 5.1 that gamma and m_dot_break 'are not statistically independent quantities, because both are derived from the same exponential baseline,' and we frame the comparison as 'exploratory.' However, we agree that Conclusion point 3 presents the non-correlation as a more substantive finding than is warranted given this shared dependence. In the revised manuscript, we will (1) add an explicit discussion of how the non-independence could in principle either suppress or inflate the correlation, depending on the covariance structure of the fitted baseline parameters; (2) note that without dedicated Monte Carlo simulations of the joint distribution of gamma and m_dot_break under the null hypothesis of a shared baseline, we cannot fully disentangle statistical artifact from physical result; and (3) revise Conclusion point 3 to restrict the claim to a more cautious statement — namely, that the data show no evidence for a strong correlation between the recession slope and the break scale, but that the non-independence of the two quantities limits the physical interpretation of this absence. We will not remove the comparison from the paper, as we believe the gamma–m_dot_break diagram remains a useful exploratory tool for visualizing the diversity of recession behavior, but we will ensure the conclusions do not overstate what the correlation test can demonstrate.","revision_made":"partial","referee_comment":"Section 5.1, Eq. (6) and Figure 3B: gamma and m_dot_break are both derived from the same exponential baseline and are not statistically independent. The reported non-correlation (r_S = 0.3, 0.6 sigma) may reflect the statistical structure of the construction rather than a physical result. The conclusion states this as a substantive finding (point 3)."}],"tokens_in":24073,"tokens_out":1858,"duration_ms":251352,"standing_objections":["The non-independence of gamma and m_dot_break (Major Comment 3) cannot be fully resolved without dedicated simulations of the joint parameter distribution under controlled null hypotheses, which is beyond the scope of the current revision. We will acknowledge this limitation explicitly but cannot eliminate it."]},"desk_editor":{"model":"glm-5.2","letter":"The main thing to know: this paper takes the luminosity-deficit method from You et al. (2023) — which was demonstrated on MAXI J1820+070 — and applies it systematically to 24 outbursts across 26 BHXRBs using RXTE/PCA archival data. That population-scale extension is the real contribution here, and it is done carefully. The spectral decomposition (thcomp⊗diskbb + Gaussian) is standard but appropriate for the PCA band, the MCMC baseline fitting with free interval boundaries is a step above manual selection, and the kTe sensitivity check (50/150/300 keV) shows the disk parameters are robust. The timing cross-check — characteristic frequencies decreasing as the inferred truncation radius grows — is qualitative but directionally correct and honestly labeled as such. The diversity analysis in the γ–ṁ_break plane is genuinely new and raises real questions about what controls the onset and pace of recession. The paper is also transparent about limitations, which matters: Section 5.2 explicitly discusses the passive-disk and outflowing-corona alternatives, and the authors call R_tr/R_ISCO a 'tracer' rather than a measurement throughout. That framing is correct and keeps the claims proportional to what the data support. The soft spot is the one the reader and stress-test both flag: Eq. 4 is valid only if the mass inflow rate continues exponentially after t_break. If Ṁ itself deviates — due to irradiation loss, outer-disk cooling, or changing mass supply — the luminosity deficit conflates geometry with mass supply, and R_tr/R_ISCO is not tracing what it claims. The irradiation-radius check (Figure 4) addresses one specific alternative (complete irradiation loss) but does not verify exponential Ṁ persistence. The stress-test suggestion is the right one: if the disk recedes while Ṁ continues exponentially, the total luminosity L_disk + L_Comp should approximately track L_exp(t). The paper shows the Comptonized component rising as the disk drops (Figures 8–9) but never explicitly checks the sum against the baseline. That check would either validate or break the central assumption, and it is the obvious next step. The circularity concern is real but I would soften it slightly: the tracer is constructed by definition from the luminosity ratio, but the timing cross-check provides an independent (if qualitative) consistency test, so the framework is not purely tautological. No code or data products are shipped, and the baseline interval selection involves some manual judgment despite the MCMC procedure, which limits reproducibility. These are minor relative to the main assumption issue. This paper is for accretion physicists working on state transitions and disk geometry. It deserves a serious referee who can push on the exponential-persistence assumption and ask for the total-luminosity check. I lean toward accept with revisions.","headline":"Population-scale disk-recession tracer for BHXRB soft-to-hard transitions; validity hinges on unverified exponential mass-inflow assumption","tokens_in":25292,"tokens_out":669,"would_cite":true,"duration_ms":129888,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"glm-5.2","headline":"Disk luminosity deficit traces thin disk recession in black hole binaries","keywords":["black hole X-ray binaries","accretion disk","disk truncation","soft-to-hard transition","luminosity deficit","disk recession","X-ray timing","QPO"],"falsifier":"If an independent measurement of the disk inner radius (e.g., from reflection spectroscopy or reverberation lags) during the soft-to-hard transition showed the disk remaining at or near the ISCO while the luminosity deficit grows, the tracer would be measuring coronal power redistribution rather than geometric disk recession.","tokens_in":24215,"feed_emoji":"🕳️","tokens_out":1475,"duration_ms":199658,"temperature":0.7,"pith_summary":"The paper proposes that when a black hole X-ray binary transitions from its soft state to its hard state, the optically thick accretion disk does not simply fade — it physically recedes, its inner edge pulling outward from the black hole. Because direct spectral measurements of the disk inner radius become unreliable during this transition (the thermal component weakens and becomes entangled with Comptonized emission), the authors offer an alternative: fit the smooth exponential decline of thermal disk luminosity during the soft state, extrapolate that trend forward in time, and treat any deficit of the observed luminosity below the extrapolation as a sign that the disk has moved outward and is releasing less gravitational energy per unit accreted mass. The central relation is R_tr/R_ISCO = L_exp/L_obs, where the ratio of expected-to-observed disk luminosity directly gives the ratio of the truncation radius to the innermost stable circular orbit. Applying this to 24 outbursts from RXTE observations of 26 black hole X-ray binaries, the authors find the pattern is ubiquitous: exponential soft-state decay followed by a systematic luminosity deficit, accompanied by growth of the Comptonized component and a shift of timing frequencies to lower values — the latter consistent with a hot inner flow expanding as the thin disk recedes. The onset and speed of recession vary widely across sources, with no single accretion-rate threshold governing when recession begins.","feed_headline":"Disk luminosity deficit traces thin disk recession in black hole binaries","feed_subtitle":"In 24 outbursts, the thermal disk drops below its soft-state exponential decay — and the deficit maps the disk's inner edge moving outward.","key_machinery":"The luminosity-deficit tracer (Equation 4): R_tr/R_ISCO = L_exp/L_obs, where L_exp is the extrapolated soft-state exponential disk luminosity and L_obs is the observed disk luminosity after the break. The exponential baseline is fit via MCMC with free interval boundaries, and the resulting ratio converts a luminosity deficit into a dimensionless truncation radius. Timing frequencies (broadband noise break frequency and QPO characteristic frequency) serve as an independent consistency check.","core_discovery":"The ratio of extrapolated-to-observed thermal disk luminosity, R_tr/R_ISCO = L_exp/L_obs, provides a model-defined tracer of disk truncation radius evolution during the soft-to-hard transition in black hole X-ray binaries. In 24 outbursts, this tracer increases rapidly after the disk luminosity breaks below the soft-state exponential baseline, and the increase is corroborated by decreasing characteristic frequencies of broadband noise and low-frequency QPOs, consistent with an expanding hot inner flow.","pith_inferences":["The tracer's validity hinges on the assumption that mass inflow continues to follow the soft-state exponential trend after the break. If the mass supply itself deviates — for instance due to outer-disk cooling reducing the inflow rate — the luminosity deficit would partially reflect a change in mass supply rather than purely a change in accretion efficiency from disk recession. The paper acknowled","The luminosity deficit could also arise without large geometric recession if a growing fraction of accretion power is dissipated in a corona or outflow rather than in the thin disk. In that case, R_tr/R_ISCO would track an effective radiative efficiency change rather than a physical inner radius, though the timing trends would still be qualitatively consistent with inner-flow expansion.","Combining the luminosity-deficit tracer with future simultaneous reflection or reverberation measurements (e.g., from NICER+NuSTAR) could break the degeneracy between true geometric recession and coronal power redistribution, testing whether the inferred R_tr corresponds to a physical disk edge."],"forward_implications":["The luminosity-deficit method can be applied to any black hole X-ray binary outburst with sufficient soft-state coverage, providing a uniform, model-light way to compare disk evolution across sources without relying on direct spectral radius measurements.","The wide diversity in break accretion rate and recession slope across outbursts suggests that disk recession is not triggered by a single universal accretion-rate threshold, pointing to additional controlling factors such as irradiation geometry, magnetic field configuration, or disk evaporation efficiency.","If the tracer is validated against independent radius estimates (e.g., from reverberation mapping or reflection modeling in the NICER era), it could become a standard diagnostic for mapping the accretion geometry evolution during state transitions.","The method could be extended to neutron star X-ray binaries to test whether disk recession occurs in systems without an event horizon, constraining the role of the inner boundary condition."],"fun_headline_variants":["Thermal luminosity deficits trace thin disk recession in black hole binaries","Mapping disk recession in X-ray binaries via luminosity deficits","Disk luminosity deficits trace the expanding hot inner flow in X-ray binaries","Luminosity deficits track disk truncation radius during black hole transitions","Tracing the receding thin disk in black hole binaries via luminosity deficits"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The method assumes that the underlying mass inflow rate continues to follow the soft-state exponential decay after the luminosity break. If the mass inflow rate itself deviates from the exponential trend — for example because outer-disk irradiation weakens or the disk drains differently — then the luminosity deficit reflects a change in mass supply rather than a change in accretion efficiency from disk recession.","fun_headline_variants_meta":{"raw":{"variants":["Thermal luminosity deficits trace thin disk recession in black hole binaries","Mapping disk recession in X-ray binaries via luminosity deficits","Disk luminosity deficits trace the expanding hot inner flow in X-ray binaries","Luminosity deficits track disk truncation radius during black hole transitions","Tracing the receding thin disk in black hole binaries via luminosity deficits"]},"model":"glm-5.2","effort":"high","cost_usd":0.0,"raw_usage":{"total_tokens":1009,"prompt_tokens":541,"completion_tokens":468,"prompt_tokens_details":null},"tokens_in":541,"tokens_out":468,"duration_ms":71059,"temperature":1.0,"reasoning_tokens":430,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-07T21:40:58.025766+00:00","model_set":{"reader":"glm-5.2"},"falsifier":"If an independent measurement of the disk inner radius (e.g., from reflection spectroscopy or reverberation lags) during the soft-to-hard transition showed the disk remaining at or near the ISCO while the luminosity deficit grows, the tracer would be measuring coronal power redistribution rather than geometric disk recession.","supporting_citations":[],"review_version":1}