KICC papers

arXiv:2509.12301v1 Announce Type: new Abstract: Multiply-imaged supernovae (SNe) provide a novel means of constraining the Hubble constant ($H_0$). Such measurements require a combination of precise models of the lensing mass distribution and an accurate estimate of the relative time delays between arrival of the multiple images. Only two multiply-imaged SNe, Refsdal and H0pe, have enabled measurements of $H_0$ thus far. Here we detail the third such measurement for SN Encore, a $z=1.95$ SNIa discovered in JWST/NIRCam imaging. We measure the time delay, perform simulations of additional microlensing and millilensing systematics, and combine with the mass models of Suyu et al. in a double-blind analysis to obtain our $H_0$ constraint. Our final time-delay measurement is $\Delta t_{1b,1a}=-39.8_{-3.3}^{+3.9}$ days, which is combined with seven lens models weighted by the likelihood of the observed multiple image positions for a result of $H_0=66.9_{-8.1}^{+11.2} \rm{km} \rm{s}^{-1}\rm{Mpc}^{-1}$. The uncertainty on this measurement could be improved significantly if template imaging is obtained. Remarkably, a sibling to SN Encore (SN "Requiem") was discovered in the same host galaxy, making the MACS J0138.0-2155 cluster the first system known to produce more than one observed multiply-imaged SN. SN Requiem has a fourth image that is expected to appear within a few years, providing an unprecedented decade-long baseline for time-delay cosmography and an opportunity for a high-precision joint estimate of $H_0$.

arXiv:2509.11175v1 Announce Type: new Abstract: The 21-cm line of hydrogen is the most promising probe of the Dark Ages and Cosmic Dawn. We combine hydrodynamical simulations with a large-scale grid in order to calculate the effect of non-linear structure formation on the large-scale 21-cm power spectrum, focusing on redshifts $z=20-40$. As the clumping effect arises from small-scale density fluctuations, it offers a unique opportunity to probe the standard cold dark matter model in a new regime and thus potentially investigate the properties of dark matter. To this end, we also study a warm dark matter $-$ like model with a Gaussian cutoff on a scale of 50 kpc. We find that clumping has a significant impact on the large-scale 21-cm power spectrum. For example, for the Dark Ages case at $z=30$ and wavenumber $k=0.05$ Mpc$^{-1}$, small-scale clustering enhances the 21-cm power spectrum by $13\%$. Once Lyman-$\alpha$ coupling kicks in due to the first stars, the 21-cm signal strengthens, and the effect of clumping grows; it suppresses the observable power spectrum at $z=20$ by a factor of two, while the cutoff model has less than half the clumping impact. The clumping effect is significantly higher than the sensitivity of the planned Square Kilometre Array (SKA) AA$^\star$ configuration, by up to a factor of 20 for standard cold dark matter, though detection will require separation from foregrounds and from astrophysical contributions to the 21-cm power spectrum.

arXiv:2509.11055v1 Announce Type: new Abstract: It is thought that the Universe went through an early period known as the Dark Ages, during which primeval density fluctuations grew to form the first luminous objects, marking the beginning of Cosmic Dawn around 100 million years after the Big Bang. The 21-cm line of hydrogen atoms is the most promising probe of these epochs, with extensive observational efforts underway. We combine hydrodynamical simulations with a large-scale grid in order to precisely calculate the effect of non-linear structure formation on the global (sky-averaged) 21-cm radio intensity. We show that it presents a potential opportunity to probe the properties of dark matter in a new regime, corresponding to a length-scale of only 150,000 light years and a mass-scale of 20 million Solar masses. This effect can in principle be detected unambiguously during the Dark Ages, where the weak signal requires an array of global signal antennae. During Cosmic Dawn, when stellar radiation boosts the signal, a single global antenna suffices, but the clumping effect must then be separated from the effect of the stars. Our findings open new avenues for testing the nature of dark matter as well as non-standard cosmological models.

arXiv:2509.10604v1 Announce Type: new Abstract: Component separation methods mitigate the cross-contamination between different extragalactic and galactic contributions to cosmic microwave background (CMB) data. This is often done by linearly combining CMB maps from different frequency channels using internal linear combination (ILC) methods. We demonstrate that deriving power spectrum estimators directly by linearly combining auto- and cross-spectra instead of maps allows us to obtain a different constrained-optimization problem that allows fewer (deprojection) constraint equations than combining at map level using the constrained ILC method. Through simulations, we show that our Spectral internal linear combination (SpILC) produces CMB power spectrum estimators with more than 7 times smaller errorbars than constrained ILC (with thermal Sunyaev-Zel'dovich and cosmic infrared background deprojections) at $\ell\gtrsim 4000$ for Simons Observatory-like observations. Spectral ILC outperforms constrained ILC methods when some modeled components are spatially uncorrelated, e.g. the primary CMB is uncorrelated with foregrounds, and the difference in performance is most significant at noise-dominated scales. More generally, our work shows that component-separated maps with foreground deprojections do not necessarily produce minimum-variance two-or-higher-point estimators.

arXiv:2509.10117v1 Announce Type: new Abstract: While the James Webb Space Telescope (JWST) now allows identifying quiescent galaxies (QGs) out to early epochs, the photometric selection of quiescent galaxy candidates (QGCs) and the derivation of key physical quantities are highly sensitive to the assumed star-formation histories (SFHs). We aim to quantify how the inclusion of JWST/MIRI data and different SFH models impacts the selection and characterisation of QGCs. We test the robustness of the physical properties inferred from the spectral energy distribution (SED) fitting, such as M*, age, star formation rate (SFR), and AV, and study how they impact the quiescence criteria of the galaxies across cosmic time. We perform SED fitting for ~13000 galaxies at z<6 from the CEERS/MIRI fields with up to 20 optical-mid infrared (MIR) broadband coverage. We implement three SFH prescriptions: flexible delayed, NonParametric, and extended Regulator. For each model, we compare results obtained with and without MIRI photometry and dust emission models. We evaluate the impact of these configurations on the number of candidate QGCs, selected based on rest UVJ colours, sSFR and main-sequence offset, and on their key physical properties such as M*, AV, and stellar ages. The number of QGCs selected varies significantly with the choice of SFH from 171 to 224 out of 13000 galaxies, depending on the model. This number increases to 222-327 when MIRI data are used (up to ~45% more QGCs). This enhancement is driven by improved constraints on dust attenuation and M*. We find a strong correlation between AV and M*, with massive galaxies (M*~10^11 M\odot) being 1.5-4.2 times more attenuated in magnitude than low-mass systems (M*~10^9 M\odot), depending on SFH. Regardless of the SFH assumption, ~13% of QGCs exhibit significant attenuation (AV > 0.5) in support of recent JWST studies challenging the notion that quiescent galaxies are uniformly dust-free.

arXiv:2509.10455v1 Announce Type: new Abstract: There is no consensus on how baryon feedback shapes the underlying matter distribution from either simulations or observations. We confront the uncertain landscape by jointly analyzing new measurements of the gas distribution around groups and clusters -- DESI+ACT kinetic Sunyaev-Zel'dovich (kSZ) effect profiles and eROSITA X-ray gas masses -- with mean halo masses characterized by galaxy-galaxy lensing. Across a wide range of halo masses ($M_{500}=10^{13-14}M_\odot$) and redshifts ($08 \sigma$ discrepant. The FLAMINGO simulation variant with the most gas expulsion, and therefore the most suppression of the matter power spectrum relative to a dark matter only simulation, provides a good description of how much gas is expelled and how far it extends; the enhanced gas depletion is achieved by more powerful but less frequent AGN outbursts. Joint kSZ, X-ray, and lensing measurements form a consistent picture of gas expulsion beyond several $R_{500}$, implying a more suppressed matter power spectrum than predicted by most recent simulations. Complementary observables and next-generation simulations are critical to understanding the physical mechanism behind this extreme gas expulsion and mapping its impact on the large-scale matter distribution.

arXiv:2504.03839v2 Announce Type: replace Abstract: We present the first results from MARTA (Measuring Abundances at high Redshift with the T$_e$ Approach), a programme leveraging ultra-deep, medium-resolution JWST/NIRSpec spectroscopy to probe the interstellar medium (ISM) of star-forming galaxies at $z \sim 2 - 3$. We report detections of one or more auroral lines, including [O III]$\lambda4363$, [O II]$\lambda\lambda7320,7330$, [S II] $\lambda4068$, and [S III] $\lambda6312$ for 16 galaxies in the sample, providing measurements of multiple ionic temperatures. We tested the validity of the T[O II]-T[O III] relation at high redshift considering a total sample of 21 objects including literature data, and obtained a shallower slope than in the low-$z$ literature. However, such a slope is consistent with low-redshift data when ultra-low metallicity objects are considered. We assessed the correlation of the T[O II]-T[O III] relationship and its scatter on different physical parameters, finding a mild correlation with the ionisation parameter and radiation field hardness, while no significant correlation with gas density. The location of high-redshift data is also consistent with the low-$z$ literature in the T[O II]-T[S II], and T[S III]-T[O III] relations, although this conclusion is limited with low-number statistics. Finally, we leveraged our sample together with a comprehensive compilation of galaxies with [O III]$\lambda4363$ detections from the literature to recalibrate classical strong-line diagnostics at high redshift. MARTA represents a key addition in this space because it provides direct metallicities at moderately high oxygen abundances (12+log(O/H) $\sim8.0-8.4$).

arXiv:2509.09626v1 Announce Type: new Abstract: We present an investigation into the quenching of simulated galaxies across cosmic time, honing in on the role played by both intrinsic and environmental mechanisms at different epochs. In anticipation of VLT-MOONRISE, the first wide-field spectroscopic galaxy survey to target cosmic noon, this work provides clear predictions to compare to the future observations. We investigate the quenching of centrals, high-mass satellites, and low-mass satellites from two cosmological hydrodynamical simulations: IllustrisTNG and EAGLE. Satellites are split according to bespoke mass thresholds, designed to separate environmental and intrinsic quenching mechanisms. To determine the best parameter for predicting quiescence, we apply a Random Forest classification analysis for each galaxy class at each epoch. The Random Forest classification determines supermassive black hole mass as the best predictor of quiescence in centrals and high-mass satellites. Alternatively, the quenching of low-mass satellites is best predicted by group halo mass, at all epochs. Additionally, we investigate the evolution in the dependence of the quenched fraction with various parameters, revealing a more complex picture. There is strong evidence for the rejuvenation of star formation from z = 2 to z = 0 in EAGLE, but not in IllustrisTNG. The starkest discrepancy between simulations rests in the mass threshold analysis. While IllustrisTNG predicts the existence of environmentally quenched satellites visible within the survey limits of MOONRISE, EAGLE does not. Hence, MOONRISE will provide critical data that is needed to evaluate current models, and constrain future models, of quenching processes.

arXiv:2502.21119v3 Announce Type: replace Abstract: Dust is a fundamental component of the interstellar medium within galaxies, as dust grains are highly efficient absorbers of ultraviolet (UV) and optical photons. Accurately quantifying this obscuration is crucial for interpreting galaxy spectral energy distributions (SEDs). The extinction curves in the Milky Way (MW) and Large Magellanic Cloud exhibit a strong feature known as the 2175 {\AA} UV bump, most often attributed to small carbonaceous dust grains. This feature was recently detected in faint galaxies out to z=7.55, suggesting rapid formation channels. Here, we report the detection of a strong UV bump in a luminous Lyman-break galaxy at z_prism=7.11235, GNWY-7379420231, through observations taken as part of the NIRSpec Wide GTO survey. We fit a dust attenuation curve that is consistent with the MW extinction curve within 1{\sigma}, in a galaxy just ~700 Myr after the Big Bang. From the integrated spectrum, we infer a young mass-weighted age (t~22-59 Myr) for this galaxy, however spatially resolved SED fitting unveils the presence of an older stellar population (t~252 Myr). Furthermore, morphological analysis provides evidence for a potential merger. The underlying older stellar population suggests the merging system could be pre-enriched, with the dust illuminated by a merger-induced starburst. Moreover, turbulence driven by stellar feedback in this bursty region may be driving polycyclic aromatic hydrocarbon formation through top-down shattering. The presence of a UV bump in GNWY-7379420231 solidifies growing evidence for the rapid evolution of dust properties within the first billion years of cosmic time

arXiv:2506.12141v2 Announce Type: replace Abstract: We report the discovery of a little red dot (LRD), dubbed BiRD ('big red dot'), at $z=2.33$ in the field around the $z=6.3$ quasar SDSSJ1030+0524. Using NIRCam images, we identified it as a bright outlier in the $F200W-F356W$ color vs $F356W$ magnitude diagram of point sources in the field. The NIRCam/WFSS spectrum reveals the emission from HeI$\lambda 10830$ and PaG line, both showing a narrow and a broad ($FWHM\gtrsim 2000\ \rm kms^{-1}$) component. The HeI line is affected by an absorption feature, tracing dense gas with HeI column density in the $2^3S$ level $N\sim 0.5-1.2\times 10^{14}\rm cm^{-2}$, depending on the location of the absorber, which is outflowing at the speed of $\Delta v \sim -830\ \rm kms^{-1}$. As observed in the majority of LRDs, BiRD does not show X-ray or radio emission. The BH mass and the bolometric luminosity, both inferred from the PaG broad component, amount to $M_{\rm BH}\sim 10^8\rm M_{\odot}$ and $L_{\rm bol}\sim 2.9\times 10^{45}\rm ergs^{-1}$, respectively. Intriguingly, BiRD presents strict analogies with other two LRDs spectroscopically confirmed at cosmic noon, GN-28074 ("Rosetta Stone") at $z=2.26$ and RUBIES-BLAGN-1 at $z=3.1$. The blueshifted HeI absorption detected in all three sources suggests that gas outflows may be common in LRDs. We derive a first estimate of the space density of LRDs at $z<3$ based on JWST data, as a function of $L_{\rm bol}$ and BH mass. The space density is only a factor of $\sim 2-3$ lower than that of UV-selected quasars with comparable $L_{\rm bol}$ and $z$, meaning that the contribution of LRDs to the broader AGN population is also relevant at cosmic noon. A similar trend is also observed in terms of BH masses. If, as suggested by recent theories, LRDs probe the very first and rapid growth of black hole seeds, our finding may suggest that the formation of black hole seeds remains efficient at least up to cosmic noon.

arXiv:2509.07064v1 Announce Type: new Abstract: The James Webb Space Telescope (\textit{JWST}) opened a new observational window on the primordial Universe. Here we present new JWST NIRSpec integral field spectroscopy (IFS) observations of the $z=7.54$ quasar ULAS J1342+0928 obtained as part of the Galaxy Assembly with NIRSpec IFS (GA-NIFS) GTO programme. The new data-set obtained with both the prism ($R\sim100$) and the high-resolution grating ($R\sim2700$) allow for a complete description of the quasar emission from the rest-frame UV to optical bands. The low-resolution data reveal the presence of [\ion{O}{iii}] emission on $\sim$7 kpc scales, well above the typical galaxy size at this redshift, likely associated with a past outflow event. Additionally, the high-resolution observations show a more energetic ionised outflow on nuclear scales ($\lesssim 0.6$ kpc). The total ionised mass outflow rate ranges between 50 and 300 $\rm M_{\odot} \, yr^{-1}$ where the significant spread is mostly due to the lack of tight constraints on the electron density. This range overlaps in part with the star formation rate range (85--545 $\rm M_{\odot} \, yr^{-1}$), implying that the nuclear outflow could ultimately lead to an early star formation quenching. By employing an accretion disc modelling, for the first time on \textit{JWST} data, we manage to robustly estimate the black hole mass and the bolometric luminosity, $\rm \log(M_{BH}/(M_{\odot}))=9.2\pm 0.2$ and $\rm \log(L_{bol}/(erg \, s^{-1}))=46.8\pm 0.1$, respectively. We derive an Eddington ratio of $\rm \lambda_{Edd}\sim 0.4$, challenging the paradigm of widespread super-Eddington accretion in quasars at the epoch of reionisation.

arXiv:2503.11457v3 Announce Type: replace Abstract: We derive Fe-abundance ratios of 7 galaxies at $z=9-12$ with $-22

arXiv:2509.05724v1 Announce Type: cross Abstract: Recent advances in neural density estimation have enabled powerful simulation-based inference (SBI) methods that can flexibly approximate Bayesian inference for intractable stochastic models. Although these methods have demonstrated reliable posterior estimation when the simulator accurately represents the underlying data generative process (GDP), recent work has shown that they perform poorly in the presence of model misspecification. This poses a significant problem for their use on real-world problems, due to simulators always misrepresenting the true DGP to a certain degree. In this paper, we introduce robust variational neural posterior estimation (RVNP), a method which addresses the problem of misspecification in amortised SBI by bridging the simulation-to-reality gap using variational inference and error modelling. We test RVNP on multiple benchmark tasks, including using real data from astronomy, and show that it can recover robust posterior inference in a data-driven manner without adopting tunable hyperparameters or priors governing the misspecification.

arXiv:2509.05459v1 Announce Type: new Abstract: The JWST has enabled the discovery of Active Galactic Nuclei at high redshifts. The intrinsic UV spectrum of GN-z11 at redshift z = 10.6 has a spectral slope compatible with a standard accretion disc. By fitting a disc model to its spectrum, we find that the mass of the black hole must be above 1.6e7 Msun in order that it lies below the Eddington limit. We define this mass as the Eddington mass of the black hole. We note that the spectral shape is consistent with that of accreting stellar mass black holes sources in their soft state, for which no variability is expected. Mom-z14 is a more distant object at z = 14.44 and has a similar UV slope. Disc model-fitting gives a similar result but lower mass accretion rate. We also examine 3 further high redshift objects: GS z14-1, GHZ2 and GS-z11-1 at z = 13.86, 12.34 and 11.28, again obtaining similar results. If sub-Eddington accretion discs are indeed the origin of much of the UV emission from these objects, then the existence of massive black holes less than 440 and 290 Myr after the Big Bang point either to exceptional black hole seeds or to primordial black holes. The observed spread of UV spectral slopes in high redshift objects suggests that our approach may be relevant to about half of that population.

arXiv:2509.05423v1 Announce Type: new Abstract: JWST has revealed a previously unknown population of low-luminosity active galactic nuclei (AGN) in the early Universe. These JWST-AGN at high redshifts are characterised by a set of peculiar properties, including unusually weak X-ray emission. Here we investigate the apparent lack of X-ray emission in the framework of the ``AGN radiative dusty feedback'' scenario based on the effective Eddington limit for dust. We analyse how the boundary in the $N_\mathrm{H} - \lambda$ plane, defined by the column density versus the Eddington ratio, is modified as a function of the dusty gas parameters (metallicity, dust grain size and composition). Low metallicity gas with little dust content tends to survive against radiation pressure, and likely accumulates in the nuclear region. We suggest that such dust-poor gas can provide long-lived absorption and may lead to heavy X-ray obscuration, as observed in early JWST-AGN. The blowout vs. stalling condition of the obscuring clouds indicates that higher metallicities are required to eject heavier column densities, while large columns of gas can stall in low metallicity environments. Therefore the metallicity may play a key role in the AGN radiative dusty feedback scenario. We discuss how other peculiar properties of JWST-AGN --such as Balmer absorption features and weak radio emission-- may be naturally interpreted within the same physical framework.

arXiv:2509.04336v1 Announce Type: cross Abstract: We present a GPU-accelerated implementation of the gravitational-wave Bayesian inference pipeline for parameter estimation and model comparison. Specifically, we implement the `acceptance-walk' sampling method, a cornerstone algorithm for gravitational-wave inference within the bilby and dynesty framework. By integrating this trusted kernel with the vectorized blackjax-ns framework, we achieve typical speedups of 20-40x for aligned spin binary black hole analyses, while recovering posteriors and evidences that are statistically identical to the original CPU implementation. This faithful re-implementation of a community-standard algorithm establishes a foundational benchmark for gravitational-wave inference. It quantifies the performance gains attributable solely to the architectural shift to GPUs, creating a vital reference against which future parallel sampling algorithms can be rigorously assessed. This allows for a clear distinction between algorithmic innovation and the inherent speedup from hardware. Our work provides a validated community tool for performing GPU-accelerated nested sampling in gravitational-wave data analyses.

arXiv:2509.03593v1 Announce Type: new Abstract: Transient events associated with supermassive black holes provide rare opportunities to study accretion and the environments of supermassive black holes. We present a multiwavelength study of AT2020adpi (ZTF20acvfraq), a luminous optical/UV transient in the nucleus of the galaxy WISEA J231853.77$-$103505.6 ($z=0.26$) that exhibits the properties of an ambiguous nuclear transient. Near peak, its spectral energy distribution is well described by a power law ($\lambda L_\lambda \propto \lambda^{-\alpha}$, $\alpha = 0.44 \pm 0.04$), with a maximum $g$-band luminosity of $(3.6 \pm 0.6)\times10^{44}$ erg s$^{-1}$, which is consistent with luminous AGN flares. We detect a strong mid-infrared flare ($L_\mathrm{peak}^{\mathrm{MIR}} = (2.3 \pm 0.05)\times10^{44}$ erg s$^{-1}$) delayed by $\sim$240 rest-frame days, indicating a hot dust echo from material at $\sim$0.2 pc. The optical and near-infrared spectra show broad H, He I, [OIII] lines, as well as narrow Fe II, and prominent Mg II, which is a combination not typical of TDEs. Taken together, these features suggest AT2020adpi is an ambiguous nuclear transient, where an accretion episode was triggered by stellar disruption of an accretion disk or instabilities within an active nucleus. This source demonstrates the need for careful multiwavelength analysis to distinguish between extreme AGN variability and TDEs.

arXiv:2509.03585v1 Announce Type: new Abstract: We aim to study the clustering of metal absorption lines and the structures that they arise in as a function of cosmic time. We focus on C IV and Si IV absorption features that are identified along a given quasar sightline. We exploit the two-point correlation function (2PCF) to investigate the clustering of these structures as a function of their separation. We utilise the E-XQR-30 data to perform a novel analysis at z>5. We also draw on literature surveys (including XQ-100) of lower redshift quasars to investigate the possible evolution of this clustering towards cosmic noon (i.e., z~2-3). We find no significant evolution with redshift when considering the separation of absorbers in velocity space. Since we are comparing data across a large interval of cosmic time, we also consider the separation between absorbers in the reference frame of physical distances. In this reference frame, we find that the amplitude of the clustering increases with cosmic time for both C IV and Si IV on the scales of <1500 physical kpc. For the first time, we assess the 2PCF of C IV and Si IV close to the epoch of reionisation utilising the absorber catalogue from the E-XQR-30 survey. We compare this with lower redshift data and find that, on small scales, the clustering of these structures grows with cosmic time. We compare these results to the clustering of galaxies in the GAEA simulations. It appears that the structures traced by C IV are broadly comparable to the galaxies from the considered simulations. The clustering is most similar to that of the galaxies with virial masses M~10^10.5 M_sun. We require tailor-made simulations to investigate the full range of factors contributing to the observed clustering. Future ground-based spectrographs will further facilitate surveys of absorbers at this epoch with increased sensitivity.

arXiv:2509.03458v1 Announce Type: new Abstract: The kinematic Sunyaev-Zel'dovich (kSZ) effect in the cosmic microwave background (CMB) is a powerful probe of gas physics and large-scale structure (LSS) in our universe. We consider the "projected-field" kSZ estimator, which involves cross-correlating a foreground-cleaned, filtered, squared CMB temperature map with an LSS tracer, and requires no individual tracer redshifts. We compare $\verb|class_sz|$ halo model calculations of projected-field kSZ cross-correlations with measurements of these signals from the Websky numerical simulations. We cross-correlate halo density maps from Websky with various CMB secondary signals. We first validate our halo model by comparing its predictions for thermal SZ (tSZ) and patchy screening ($\tau$) cross-correlations to measurements of these signals from Websky. We consider three different halo redshift ranges in our comparisons. We also construct our own kSZ, tSZ, and $\tau$ maps to validate the form of the relevant profiles. Following the tSZ and $\tau$ validation, we compare projected-field kSZ calculations between the halo model and the simulations. We use filters constructed for $\textit{Planck}$ and the Simons Observatory (SO) to assess the accuracy of the halo-model kSZ predictions for experiments of differing sensitivity. Overall, we find good agreement, particularly at $\textit{Planck}$ sensitivity. However, we find an $\approx$ 20$\%$ difference between our halo model and the simulations for SO, which significantly exceeds the predicted error bars. We note that our halo model includes only the dominant expected term in the projected-field kSZ signal; the magnitude of the difference between our model and the simulations is consistent with previous predictions for terms arising from other contractions in the theory calculation. These terms will need to be included to obtain unbiased inference from upcoming projected-field kSZ measurements.

arXiv:2509.02769v1 Announce Type: new Abstract: We present a detailed multi-wavelength afterglow study of the short GRB 250704B, extensively monitored in optical and near-infrared bands. Its afterglow displays an unusually long-duration plateau followed by an achromatic break and a steep decline, deviating from canonical GRB afterglows. While long plateaus are often explained by central engine activity, we find that for GRB 250704B, an energy injection model requires unreasonable parameters. The afterglow is better explained by an off-axis power-law structured jet with a narrow core ($\theta_c \approx 0.7^{\circ}$) viewed at a modest angle ($\theta_v \approx 1.9^{\circ}$). A comparison with GRB 170817A shows that both events are consistent with the off-axis structured jet scenario, where the shape of the light curve is governed primarily by the geometry of the jet and the viewing angle rather than the energetics, microphysical parameters, or external density. Our results underscore the importance of incorporating the jet structure in GRB modeling.