{"id":"60fa9e2c-5926-49a1-9994-2326051cb6ba","arxiv_id":"2607.14696","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"TDE 2024aepd exhibits an early near-infrared excess consistent with free-free emission from a reprocessing outflow, and a rapid disk-to-corona X-ray transition.","lead":"A tidal disruption event more than a billion light-years away shows an early near-infrared glow and a rapid shift in X-ray emission from a disk to a hot corona. The infrared glow is best explained by a dense outflow reprocessing the disk light—a feature now seen in three such events.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Claimed monotonic rise of the NIR break is not supported by Table 1: νbreak decreases between the first two epochs, and the epoch-3 value relies on a non-contemporaneous K-band point.","rationale":"The reader's CONDITIONAL verdict is appropriate, but the weakest assumption identified by the reader was the homogeneous/spherical free-free model geometry. My stress-test finds a more specific and directly checkable problem in the paper's own Table 1: the claimed monotonic increase of νbreak is not visible in the quoted best-fit values, which decrease from epoch 1 to epoch 2 and only increase at epoch 3 with the help of a borrowed K-band point. This is an internal inconsistency, not just a model-dependence issue. It weakens the evolutionary conclusion but does not invalidate the central NIR-excess detection or the X-ray state-transition claim. The paper should be accepted conditionally, with the abstract and Section 4.1 revised to describe the break evolution more cautiously. I therefore leave the reader's verdict unchanged.","tokens_in":27064,"tokens_out":22502,"duration_ms":226827,"concrete_test":"Re-fit the three combined UV–optical–NIR SEDs using only contemporaneous NIR detections, replacing the epoch-3 borrowed K point with the actual K-band upper limit or fitting K_3 as a free parameter. Then bootstrap the MCMC fits and compute the posterior probability that νbreak(41d) < νbreak(68d) < νbreak(101d). If this probability is not high, revise the abstract and Section 4.1 to state that νbreak is consistent with no monotonic evolution, rather than claiming it systematically shifts to higher frequencies.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract and Section 4.1 assert that the UV–optical-to-NIR break shifts to higher frequencies while the density-profile index remains nearly constant. Table 1 does not support this. At +41, +68, and +101 days, the best-fit break frequencies are νbreak = 2.89, 2.63, and 3.67 ×10^14 Hz, respectively. Thus the first two epochs show a decrease, not an increase. Moreover, the epoch-3 value is not an independent measurement: the source was not detected in K at +101 d, so the epoch-2 K-band measurement was adopted for epoch 3. With only J and K bands per epoch, the free-free model's parameters are already weakly constrained, and the epoch-3 increase rests on this borrowed datum. The claim of 'systematically increases' is therefore contradicted by the paper's own best-fit values, undermining the conclusion of evolving reprocessing conditions within a broadly unchanged density structure. The NIR excess itself and the free-free versus dust-echo preference may still stand, but the temporal-evolution narrative is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents multi-wavelength (radio, NIR, optical/UV, X-ray) observations of the tidal disruption event 2024aepd over the first ~300 days after discovery. It reports an early near-infrared excess above the UV-optical blackbody, with an approximately flat NIR power-law spectrum, and argues on the basis of a free-free reprocessing model that this excess is more plausibly explained by a photospheric reprocessing envelope than by a conventional dust echo. The X-ray data show an initially thermal-dominated spectrum with a hard tail, transitioning to a power-law-dominated and progressively harder spectrum around day 178, which the authors interpret as the emergence and strengthening of a corona. The paper also compiles a sample of TDEs with X-ray hard excesses and reports a positive correlation between photon index and Eddington ratio. The authors identify 2024aepd as the third TDE with an early-time NIR excess.","tokens_in":27379,"tokens_out":5551,"duration_ms":63519,"significance":"If the free-free interpretation is correct, the paper adds a valuable new data point to a very small sample of early NIR excesses in TDEs and supports the idea that frequency-dependent reprocessing can be diagnosed in the NIR. The X-ray spectral state transition is well documented and consistent with the growing population of TDEs showing disk-to-corona transitions. The paper's strengths include broad multi-wavelength coverage, careful host-galaxy subtraction, standard processing of Swift/XRT and ground-based data, explicit model-comparison tables, and an unusually candid set of caveats about the spherical, homogeneous assumptions of the free-free model and the tension between the inferred envelope mass and the MOSFiT stellar mass. However, the paper's central temporal-evolution claim—that the UV-optical-to-NIR break frequency increases while the density index stays constant—is not supported by the paper's own best-fit values in Table 1, and the free-free fits are underconstrained by the available two-band NIR data. The paper is publishable after the evolutionary narrative is either properly supported or substantially softened.","major_comments":[{"comment":"The abstract and §4.1 state that νbreak 'systematically increases' with time while s remains nearly constant. Table 1 gives νbreak = 2.89(+0.24/−0.33), 2.63(+0.27/−0.37), and 3.67(+0.76/−0.44) ×10^14 Hz for the +41, +68, and +101 d epochs. The first two epochs therefore show a decrease, not an increase, and the uncertainties overlap substantially. Moreover, the epoch-3 value is not an independent measurement: the source was undetected in K at +101 d, so the epoch-2 K-band datum was adopted. With only J and K per epoch, the claim of a systematic break evolution is not established. This claim is load-bearing for the conclusion of 'evolving reprocessing conditions within a broadly unchanged density structure.' The authors should either remove the monotonic-increase claim, or re-fit epoch 3 with only the J-band constraint and show whether any break evolution is statistically required, and re","section":null},{"comment":"The conclusion that the density-profile index s remains nearly constant is essentially a reparameterization of the fitted NIR spectral slope: in the adopted model, α_IR = (4s−6)/(3s−2), and s is a free parameter fit to two NIR bands per epoch. The statement that 's remains consistent within the uncertainties' is therefore not an independent test of structural stability. I recommend presenting the joint posterior of (s, νbreak), or performing a model comparison between (i) a single s with freely evolving νbreak and (ii) independent s at each epoch. Without such a test, the stable-density-structure conclusion is weaker than the text implies. This also bears on the comparison with AT2019azh in Fig. 8.","section":null},{"comment":"The argument against a dust echo relies on taking the first NIR observation as an approximate upper limit on the IR peak and deriving τ < 17 days, compared with the sublimation-radius delay of ~743 days. Section 3.2 explicitly notes that the NIR observations 'only sample the declining phase,' so the first NIR epoch does not directly constrain the time of the NIR peak. Although an upper limit on the delay can still be derived if the source was already declining at +41 d, this should be stated as a model-dependent assumption. The later discussion of anisotropic emission and line-of-sight geometry already acknowledges the main escape route for a dust echo; the quantitative τ < 17 d statement overstates the constraint and should be softened.","section":null}],"minor_comments":[{"comment":"Typos: 'primaly' should be 'primarily' and 'primally' should be 'primarily'; 'efficiency' and 'Office' contain non-standard ligature/encoding artifacts that should be corrected in the journal production step.","section":null},{"comment":"'DESI is amounted on the 4-meter Mayall Telescope' should read 'mounted'.","section":null},{"comment":"The caveat that the K-band template image could contain a late-rising echo, which would cause the true K fluxes to be underestimated, is stated in §2.4 but is not repeated when the K-band points are used to fit the free-free model. This caveat is relevant to the light-travel-time argument and should be recalled in §4.1.","section":null},{"comment":"The header 'Epoch 2K' for the epoch-3 NIR observations is confusing; it should read 'Epoch-2 K-band' with an explicit note that this datum is non-contemporaneous with the epoch-3 J band.","section":null},{"comment":"Several arXiv identifiers appear malformed (e.g., arXiv:2604.160934, arXiv:2602.21624); please check the journal's reference formatting requirements.","section":null}],"recommendation":"major_revision","confidential_remarks":"The main issue is the unsupported monotonic νbreak trend. If the authors rephrase the conclusion to state that the NIR spectral slope is consistent with a constant density index, while noting that the break-frequency evolution is not robustly measured given the two-band data and the borrowed epoch-2 K point, the paper would be acceptable. The Γ–LX correlation is interesting but should be framed as an exploratory compilation given the selection criteria."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid observational paper on a new TDE. The early NIR excess is real—two epochs of J and K clearly sit above the UV/optical blackbody tail, and the flat spectrum (αIR ≈ 0) is a reasonable match to free-free reprocessing. The X-ray data also convincingly show a transition from disk-dominated to power-law-dominated around day 178, with the photon index dropping from ~3 to ~1.8. That part I'd trust. The paper is worth reading for those two observational results.\n\nThe soft spots are in the interpretation. The claim in the abstract that 'the UV-optical-to-NIR break shifts to higher frequencies as the density-profile index remains nearly constant' is not supported by Table 1. The best-fit break frequencies are 2.89, 2.63, and 3.67 ×10^14 Hz across the three epochs. Between epochs 1 and 2 it goes down, not up; the third point uses the K-band measurement from epoch 2 because the source wasn't detected in K at epoch 3. So the 'systematic increase' narrative rests on borrowed data and a change that is not monotonic in the actual best-fit values. I'd also flag the circularity issue the reader noted: the density slope s is a fitted parameter in the free-free model, so saying s is 'nearly constant' is mostly restating that the two NIR bands give consistent slopes, not an independent confirmation of a stable density structure. The dust-echo argument is more convincing—the inferred sublimation radius implies a lag of ~700 days, much longer than the observed <17-day constraint—but the authors themselves leave dust echo open because of anisotropic geometries.\n\nThe Γ-LX correlation is suggestive but not a headline result. It is a positive correlation across a heterogeneous sample with ρ=0.46, p=0.0006, but the optically-selected subsample drops to ρ=0.31, p=0.05. The sample selection may bias against low-luminosity sources, so I wouldn't lean on this.\n\nOverall: this is a careful data paper with a likely real NIR excess and a clean X-ray state transition. The free-free model is plausible but not uniquely constrained, and the temporal evolution claim is overreaching. My advice: send it to a referee, but the referee should ask the authors to rewrite the break-evolution paragraph, show the epoch-3 fit without the borrowed K point, and soften the abstract.","headline":"Solid TDE data paper with a likely real NIR excess and clean X-ray state transition, but the break-evolution claim is contradicted by the paper's own table.","tokens_in":28041,"tokens_out":3778,"would_cite":true,"duration_ms":34136,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A tidal disruption event reveals an early near-infrared excess best explained by free-free emission from a reprocessing envelope, plus a disk-to-corona X-ray transition near day 178.","keywords":["tidal disruption event","near-infrared excess","free-free emission","reprocessing envelope","X-ray corona","spectral state transition","accretion disk","supermassive black hole"],"falsifier":"A single NIR spectrum with emission or absorption features characteristic of free-free processes would discriminate directly; more practically, observing the NIR excess before day 40 and tracking its rise would test the claimed <17-day lag. A dust-echo model predicts a delayed, roughly 1100 K blackbody with a light-travel lag comparable to the sublimation radius, whereas free-free predicts an immediate, flat power law that tracks the UV-optical decline.","tokens_in":26960,"feed_emoji":"🔭","tokens_out":3868,"duration_ms":37662,"temperature":0.7,"pith_summary":"The paper follows the tidal disruption event 2024aepd for roughly 300 days across radio, near-infrared, optical, ultraviolet, and X-ray bands. It tries to establish two things: the early near-infrared excess, with a nearly flat spectrum, is more plausibly free-free emission from a reprocessing photospheric envelope than a dust echo; and the X-ray spectrum switches from a thermal, disk-dominated state to a power-law, corona-dominated state around day 178. If right, this event joins two earlier ones as the third known TDE with an early-time NIR excess, and it gives a rare, time-resolved view of corona formation around a supermassive black hole.","feed_headline":"Third TDE shows early infrared excess from gas, not dust","feed_subtitle":"Near-infrared light points to reprocessing gas, and X-rays reveal a disk-to-corona switch in real time.","key_machinery":"The central model is frequency-dependent free-free reprocessing in a homogeneous, spherical, electron-scattering-dominated envelope with a power-law density profile rho ∝ r^{-s}. Because free-free opacity rises toward low frequencies, NIR photons thermalize at larger radii than optical/UV photons, boosting the NIR flux and producing a power-law spectrum shallower than the Rayleigh-Jeans tail; the break frequency connects the blackbody and power-law branches and, with the fitted slope s, yields the thermalization radius, density normalization, and enclosed mass. The X-ray analysis uses a two-component spectral decomposition (thermal disk plus power law) to track the state transition.","core_discovery":"For TDE 2024aepd, the near-infrared excess appears by day ~40 with a power-law slope consistent with zero, far flatter than the Rayleigh-Jeans tail of the UV-optical blackbody. Because the light-travel time to the inferred dust sublimation radius (~740 days) far exceeds the observed lag (<17 days), the authors argue a conventional dust echo is disfavoured and propose instead that the excess comes from free-free emission in a dense, extended reprocessing layer. In the same source, the X-ray spectrum, initially a thermal disk plus a hard excess, becomes purely power-law by ~178 days and hardens as the disk fades, which they interpret as the rapid emergence and strengthening of a corona. They a","pith_inferences":["If the free-free interpretation is right, the NIR excess directly probes gas mass and density profile; targeted NIR spectroscopy might look for the expected frequency dependence or polarization to confirm the mechanism.","The dust-echo alternative could be tested by measuring the NIR rise before day 40 and by searching for a late-rising echo; a detection of a delayed, ~1100 K component would overturn the free-free conclusion.","The same chromatic-reprocessing framework may apply to fast blue optical transients and to the V-shaped SEDs of Little Red Dots, making early-time NIR monitoring a useful discriminant in those populations."],"forward_implications":["Early-time NIR coverage may reveal that a substantial fraction of TDEs show such excesses, and the flat NIR slope can be used as a diagnostic of the reprocessing layer's density structure.","The measured break-frequency shift with roughly constant density slope implies the reprocessing envelope's structure persists while its density declines, constraining outflow and reprocessing models.","The disk-to-corona transition at roughly 0.5% Eddington supports using TDEs as fast laboratories for accretion-state transitions, analogous to X-ray binaries.","The positive photon-index versus X-ray-luminosity correlation in the TDE sample, without the low-luminosity anti-correlation branch, points to a lack of very low-Eddington observations rather than a fundamentally different accretion flow."],"fun_headline_variants":["TDE 2024aepd: early IR excess signals gas, not dust","TDE 2024aepd: gas, not dust, explains early infrared excess","Early IR excess in TDE 2024aepd likely from gas, not dust","Third TDE with early IR excess tied to reprocessing gas"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The free-free conclusion rests on assuming the reprocessing material is a homogeneous, spherical, electron-scattering-dominated layer with a power-law density profile; if the outflow is clumpy or anisotropic, the inferred masses and the claim of a stable density structure could fail, and the paper itself notes a dust echo cannot be fully excluded.","fun_headline_variants_meta":{"raw":{"variants":["TDE 2024aepd: early IR excess signals gas, not dust","TDE 2024aepd: gas, not dust, explains early infrared excess","Early IR excess in TDE 2024aepd likely from gas, not dust","Third TDE with early IR excess tied to reprocessing gas"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000726,"raw_usage":{"total_tokens":3107,"prompt_tokens":779,"completion_tokens":2328,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":523,"completion_tokens_details":{"reasoning_tokens":2240}},"tokens_in":523,"tokens_out":2328,"duration_ms":16045,"temperature":1.0,"reasoning_tokens":2240,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T01:18:36.607592+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A single NIR spectrum with emission or absorption features characteristic of free-free processes would discriminate directly; more practically, observing the NIR excess before day 40 and tracking its rise would test the claimed <17-day lag. A dust-echo model predicts a delayed, roughly 1100 K blackbody with a light-travel lag comparable to the sublimation radius, whereas free-free predicts an immediate, flat power law that tracks the UV-optical decline.","supporting_citations":[],"review_version":1}