Search

We present first results from James Webb Space Telescope (JWST) Near-Infrared Spectrograph (NIRSpec), Mid-Infrared Instrument (MIRI), and Keck Cosmic Webb Imager (KCWI) integral field spectroscopy of the powerful but highly obscured host-galaxy of the jetted radio source Cygnus A. We detect 169 infrared emission lines at 1.7--27 micron and explore the kinematics and physical properties of the extended narrow-line region (NLR) in unprecedented detail. The density-stratified NLR appears to be shaped by the initial blow-out and ongoing interaction of the radio jet with the interstellar medium, creating a multi-phase bicone with a layered structure composed of molecular and ionized gas. The NLR spectrum, with strong coronal emission at kpc-scale, is well-modeled by AGN photoionization. We find evidence that the NLR is rotating around the radio axis, perhaps mediated by magnetic fields and driven by angular momentum transfer from the radio jet. The overall velocity field of the NLR is well described by 250 km/s outflow along biconical spiral flow lines, combining both rotation and outflow signatures. There is particularly bright [Fe II] 1.644 micron emission from a dense, high-velocity dispersion, photoionized clump of clouds found near the projected radio axis. Outflows of 600--2000 km/s are found in bullets and streamers of ionized gas that may be ablated by the radio jet from these clouds, driving a local outflow rate of 40 Msun/yr., Comment: Submitted to Astrophysical Journal

Video personalization, which generates customized videos using reference images, has gained significant attention. However, prior methods typically focus on single-concept personalization, limiting broader applications that require multi-concept integration. Attempts to extend these models to multiple concepts often lead to identity blending, which results in composite characters with fused attributes from multiple sources. This challenge arises due to the lack of a mechanism to link each concept with its specific reference image. We address this with anchored prompts, which embed image anchors as unique tokens within text prompts, guiding accurate referencing during generation. Additionally, we introduce concept embeddings to encode the order of reference images. Our approach, Movie Weaver, seamlessly weaves multiple concepts-including face, body, and animal images-into one video, allowing flexible combinations in a single model. The evaluation shows that Movie Weaver outperforms existing methods for multi-concept video personalization in identity preservation and overall quality., Comment: Project page: https://jeff-liangf.github.io/projects/movieweaver/

Customization of text-to-image models enables users to insert custom concepts and generate the concepts in unseen settings. Existing methods either rely on costly test-time optimization or train encoders on single-image training datasets without multi-image supervision, leading to worse image quality. We propose a simple approach that addresses both limitations. We first leverage existing text-to-image models and 3D datasets to create a high-quality Synthetic Customization Dataset (SynCD) consisting of multiple images of the same object in different lighting, backgrounds, and poses. We then propose a new encoder architecture based on shared attention mechanisms that better incorporate fine-grained visual details from input images. Finally, we propose a new inference technique that mitigates overexposure issues during inference by normalizing the text and image guidance vectors. Through extensive experiments, we show that our model, trained on the synthetic dataset with the proposed encoder and inference algorithm, outperforms existing tuning-free methods on standard customization benchmarks., Comment: Project webpage: https://www.cs.cmu.edu/~syncd-project/

Text-Image-to-Video (TI2V) generation aims to generate a video from an image following a text description, which is also referred to as text-guided image animation. Most existing methods struggle to generate videos that align well with the text prompts, particularly when motion is specified. To overcome this limitation, we introduce MotiF, a simple yet effective approach that directs the model's learning to the regions with more motion, thereby improving the text alignment and motion generation. We use optical flow to generate a motion heatmap and weight the loss according to the intensity of the motion. This modified objective leads to noticeable improvements and complements existing methods that utilize motion priors as model inputs. Additionally, due to the lack of a diverse benchmark for evaluating TI2V generation, we propose TI2V Bench, a dataset consists of 320 image-text pairs for robust evaluation. We present a human evaluation protocol that asks the annotators to select an overall preference between two videos followed by their justifications. Through a comprehensive evaluation on TI2V Bench, MotiF outperforms nine open-sourced models, achieving an average preference of 72%. The TI2V Bench is released in https://wang-sj16.github.io/motif/., Comment: TI2V Bench is released in https://wang-sj16.github.io/motif/

With the recent advancements in artificial intelligence, researchers and industries are deploying gigantic models trained on billions of samples. While training these models consumes a huge amount of energy, human brains produce similar outputs (along with other capabilities) with massively lower data and energy requirements. For this reason, more researchers are increasingly considering alternatives. One of these alternatives is known as synthetic biological intelligence, which involves training \textit{in vitro} neurons for goal-directed tasks. This multidisciplinary field requires knowledge of tissue engineering, bio-materials, digital signal processing, computer programming, neuroscience, and even artificial intelligence. The multidisciplinary requirements make starting synthetic biological intelligence research highly non-trivial and time-consuming. Generally, most labs either specialize in the biological aspects or the computational ones. Here, we propose how a lab focusing on computational aspects, including machine learning and device interfacing, can start working on synthetic biological intelligence, including organoid intelligence. We will also discuss computational aspects, which can be helpful for labs that focus on biological research. To facilitate synthetic biological intelligence research, we will describe such a general process step by step, including risks and precautions that could lead to substantial delay or additional cost.

We employ real-space variational quantum Monte Carlo methods with resonating valence bond many-body wave functions to investigate electron correlation effects in the Kagome system $Zn_xCu_{4-x}O_6$. Using three trial wave functions of the Slater-Jastrow type, where (i) only the Jastrow correlation factor is optimized and the orbitals obtained by density functional theory, (ii) both the Jastrow factor and the Slater determinant are optimized, and (iii) additionally the Slater determinant is substituted by an antisymmetrized-geminal power wave function, we analyze static and dynamic correlation energies across concentrations $x=0, \frac{1}{3}, \frac{2}{3}, 1$. Our results show that the correlation energy increases with the concentration of $Zn_x$. Optimizing the Slater determinant significantly enhances the correlation energy by approximately $\sim -73(4) mHa$ per electron. Eventually, the emergence of a quantum spin liquid state driven by a long-range correlation energy is discussed.

Quantum phenomena emerging from the interaction of light and matter in low-dimensional systems hold great potential for future quantum technologies. Here, using first-principles calculations incorporating non-local van der Waals interactions and Hubbard corrections, we report simultaneous structural expansion and magnetic collapse in bilayer FeSe induced by photoexcited hot electrons. Our calculations reveal that, while bulk FeSe is paramagnetic, as observed experimentally, double-layer FeSe exhibits robust {\it staggered} antiferromagnetic order at low temperatures with a net site magnetization of $\sim 2.75~\mu_B$/Fe. However, increasing the density of photoexcited electrons systematically enhances the internal electronic entropy, leading to a complete collapse of antiferromagnetic order accompanied by an abrupt expansion of the interlayer separation. Our findings suggest the structural and magnetic properties of FeSe thin films can be finely tuned via ultrafast laser excitation, offering a pathway to control quantum phases in iron-based compounds through electronic temperature.

We present Movie Gen, a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio. We also show additional capabilities such as precise instruction-based video editing and generation of personalized videos based on a user's image. Our models set a new state-of-the-art on multiple tasks: text-to-video synthesis, video personalization, video editing, video-to-audio generation, and text-to-audio generation. Our largest video generation model is a 30B parameter transformer trained with a maximum context length of 73K video tokens, corresponding to a generated video of 16 seconds at 16 frames-per-second. We show multiple technical innovations and simplifications on the architecture, latent spaces, training objectives and recipes, data curation, evaluation protocols, parallelization techniques, and inference optimizations that allow us to reap the benefits of scaling pre-training data, model size, and training compute for training large scale media generation models. We hope this paper helps the research community to accelerate progress and innovation in media generation models. All videos from this paper are available at https://go.fb.me/MovieGenResearchVideos.

Portfolio construction has been a long-standing topic of research in finance. The computational complexity and the time taken both increase rapidly with the number of investments in the portfolio. It becomes difficult, even impossible for classic computers to solve. Quantum computing is a new way of computing which takes advantage of quantum superposition and entanglement. It changes how such problems are approached and is not constrained by some of the classic computational complexity. Studies have shown that quantum computing can offer significant advantages over classical computing in many fields. The application of quantum computing has been constrained by the unavailability of actual quantum computers. In the past decade, there has been the rapid development of the large-scale quantum computer. However, software development for quantum computing is slow in many fields. In our study, we apply quantum computing to a multi-asset portfolio simulation. The simulation is based on historic data, covariance, and expected returns, all calculated using quantum computing. Although technically a solvable problem for classical computing, we believe the software development is important to the future application of quantum computing in finance. We conducted this study through simulation of a quantum computer and the use of Rensselaer Polytechnic Institute's IBM quantum computer.

The growing volume of data produced by large astronomical surveys necessitates the development of efficient analysis techniques capable of effectively managing high-dimensional datasets. This study addresses this need by demonstrating some applications of manifold learning and dimensionality reduction techniques, specifically the Self-Organizing Map (SOM), on the optical+NIR SED space of galaxies, with a focus on sample comparison, selection biases, and predictive power using a small subset. To this end, we utilize a large photometric sample from the five CANDELS fields and a subset with spectroscopic measurements from the KECK MOSDEF survey in two redshift bins at $z\sim1.5$ and $z\sim2.2$. We trained SOM with the photometric data and mapped the spectroscopic data onto it as our study case. We found that MOSDEF targets do not cover all SED shapes existing in the SOM. Our findings reveal that Active Galactic Nuclei (AGN) within the MOSDEF sample are mapped onto the more massive regions of the SOM, confirming previous studies and known selection biases towards higher-mass, less dusty galaxies. Furthermore, SOM were utilized to map measured spectroscopic features, examining the relationship between metallicity variations and galaxy mass. Our analysis confirmed that more massive galaxies exhibit lower [OIII]/H$\beta$ and [OIII]/[OII] ratios and higher H$\alpha$/H$\beta$ ratios, consistent with the known mass-metallicity relation. These findings highlight the effectiveness of SOM in analyzing and visualizing complex, multi-dimensional datasets, emphasizing their potential in data-driven astronomical studies.

Predicting molecular impact on cellular function is a core challenge in therapeutic design. Phenomic experiments, designed to capture cellular morphology, utilize microscopy based techniques and demonstrate a high throughput solution for uncovering molecular impact on the cell. In this work, we learn a joint latent space between molecular structures and microscopy phenomic experiments, aligning paired samples with contrastive learning. Specifically, we study the problem ofContrastive PhenoMolecular Retrieval, which consists of zero-shot molecular structure identification conditioned on phenomic experiments. We assess challenges in multi-modal learning of phenomics and molecular modalities such as experimental batch effect, inactive molecule perturbations, and encoding perturbation concentration. We demonstrate improved multi-modal learner retrieval through (1) a uni-modal pre-trained phenomics model, (2) a novel inter sample similarity aware loss, and (3) models conditioned on a representation of molecular concentration. Following this recipe, we propose MolPhenix, a molecular phenomics model. MolPhenix leverages a pre-trained phenomics model to demonstrate significant performance gains across perturbation concentrations, molecular scaffolds, and activity thresholds. In particular, we demonstrate an 8.1x improvement in zero shot molecular retrieval of active molecules over the previous state-of-the-art, reaching 77.33% in top-1% accuracy. These results open the door for machine learning to be applied in virtual phenomics screening, which can significantly benefit drug discovery applications.

We present measurements on Fe$_2$O$_3$ amorphization and melt under laser-driven shock compression up to 209(10) GPa via time-resolved in situ x-ray diffraction. At 122(3) GPa, a diffuse signal is observed indicating the presence of a non-crystalline phase. Structure factors have been extracted up to 182(6) GPa showing the presence of two well-defined peaks. A rapid change in the intensity ratio of the two peaks is identified between 145(10) and 151(10) GPa, indicative of a phase change. Present DFT+$U$ calculations of temperatures along Fe$_2$O$_3$ Hugoniot are in agreement with SESAME 7440 and indicate relatively low temperatures, below 2000 K, up to 150 GPa. The non-crystalline diffuse scattering is thus consistent with the - as yet unreported - shock amorphization of Fe$_2$O$_3$ between 122(3) and 145(10) GPa, followed by an amorphous-to-liquid transition above 151(10) GPa. Upon release, a non-crystalline phase is observed alongside crystalline $\alpha$-Fe$_2$O$_3$. The extracted structure factor and pair distribution function of this release phase resemble those reported for Fe$_2$O$_3$ melt at ambient pressure., Comment: 11 pages, 4 figures, under review

The 3C 220.3 system is a rare case of a foreground narrow-line radio galaxy ("galaxy A," $z_A = 0.6850$) lensing a background submillimeter galaxy ($z_{{\rm SMG1}} = 2.221$). New spectra from MMT/Binospec confirm that the companion galaxy ("galaxy B") is part of the lensing system with $z_B = 0.6835$. New three-color HST data reveal a full Einstein ring and allow a more precise lens model. The new HST images also reveal extended emission around galaxy A, and the spectra show extended [OII] emission with irregular morphology and complex velocity structure. All indications are that the two lensing galaxies are a gravitationally interacting pair. Strong [OII] emission from both galaxies A and B suggests current star formation, which could be a consequence of the interaction. This would indicate a younger stellar population than previously assumed and imply smaller stellar masses for the same luminosity. The improved lens model and expanded spectral energy distributions have enabled better stellar-mass estimates for the foreground galaxies. The resulting dark matter fractions are ~0.8, higher than previously calculated. Deeper Chandra imaging shows extended X-ray emission but no evidence for a point X-ray source associated with either galaxy. The detection of X-rays from the radio lobes of 3C 220.3 allows an estimate of ~3 nT for the magnetic fields in the lobes, a factor of ~3 below the equipartition fields, as typical for radio galaxies., Comment: 23 pages, 10 figures, 10 tables. Updated to published version

The Chandra Source Catalog (CSC) is a virtual X-ray astrophysics facility that enables both detailed individual source studies and statistical studies of large samples of X-ray sources detected in ACIS and HRC-I imaging observations obtained by the Chandra X-ray Observatory. The catalog provides carefully-curated, high-quality, and uniformly calibrated and analyzed tabulated positional, spatial, photometric, spectral, and temporal source properties, as well as science-ready X-ray data products. The latter includes multiple types of source- and field-based FITS format products that can be used as a basis for further research, significantly simplifying followup analysis of scientifically meaningful source samples. We discuss in detail the algorithms used for the CSC Release 2 Series, including CSC 2.0, which includes 317,167 unique X-ray sources on the sky identified in observations released publicly through the end of 2014, and CSC 2.1, which adds Chandra data released through the end of 2021 and expands the catalog to 407,806 sources. Besides adding more recent observations, the CSC Release 2 Series includes multiple algorithmic enhancements that provide significant improvements over earlier releases. The compact source sensitivity limit for most observations is ~5 photons over most of the field of view, which is ~2x fainter than Release 1, achieved by co-adding observations and using an optimized source detection approach. A Bayesian X-ray aperture photometry code produces robust fluxes even in crowded fields and for low count sources. The current release, CSC 2.1, is tied to the Gaia-CRF3 astrometric reference frame for the best sky positions for catalog sources., Comment: 66 pages, 17 figures, 16 tables, accepted for publication in The Astrophysical Journal Supplement Series

We present two methods for computing the dynamic structure factor for warm dense hydrogen without invoking either the Born-Oppenheimer approximation or the Chihara decomposition, by employing a wave-packet description that resolves the electron dynamics during ion evolution. First, a semiclassical method is discussed, which is corrected based on known quantum constraints, and second, a direct computation of the density response function within the molecular dynamics. The wave packet models are compared to PIMC and DFT-MD for the static and low-frequency behaviour. For the high-frequency behaviour the models recover the expected behaviour in the limits of small and large momentum transfers and show the characteristic flattening of the plasmon dispersion for intermediate momentum transfers due to interactions, in agreement with commonly used models for x-ray Thomson scattering. By modelling the electrons and ions on an equal footing, both the ion and free electron part of the spectrum can now be treated within a single framework where we simultaneously resolve the ion-acoustic and plasmon mode, with a self-consistent description of collisions and screening., Comment: 17 pages, 9 figures

This white paper discusses the breadth of science related to active galactic nuclei (AGN) and associated phenomena to be enabled by a mission with microcalorimeter energy resolution in the soft X-ray band, a large collecting area, and wide-field imaging. Such a mission, the Line Emission Mapper (LEM), has been proposed to NASA's 2023 Astrophysics Probe Explorer call. While the science pillars of the PI-led part of the mission focus on galaxy evolution, the PI-led LEM All-Sky Survey (LASS) and General Observer/Investigator opportunities will enable vital discoveries for AGN science in the critical soft X-ray band., Comment: 22 pages, 16 figures

Sexism in online content is a pervasive issue that necessitates effective classification techniques to mitigate its harmful impact. Online platforms often have sexist comments and posts that create a hostile environment, especially for women and minority groups. This content not only spreads harmful stereotypes but also causes emotional harm. Reliable methods are essential to find and remove sexist content, making online spaces safer and more welcoming. Therefore, the sEXism Identification in Social neTworks (EXIST) challenge addresses this issue at CLEF 2024. This study aims to improve sexism identification in bilingual contexts (English and Spanish) by leveraging natural language processing models. The tasks are to determine whether a text is sexist and what the source intention behind it is. We fine-tuned the XLM-RoBERTa model and separately used GPT-3.5 with few-shot learning prompts to classify sexist content. The XLM-RoBERTa model exhibited robust performance in handling complex linguistic structures, while GPT-3.5's few-shot learning capability allowed for rapid adaptation to new data with minimal labeled examples. Our approach using XLM-RoBERTa achieved 4th place in the soft-soft evaluation of Task 1 (sexism identification). For Task 2 (source intention), we achieved 2nd place in the soft-soft evaluation.

The smart grid represents a pivotal innovation in modernizing the electricity sector, offering an intelligent, digitalized energy network capable of optimizing energy delivery from source to consumer. It hence represents the backbone of the energy sector of a nation. Due to its central role, the availability of the smart grid is paramount and is hence necessary to have in-depth control of its operations and safety. To this aim, researchers developed multiple solutions to assess the smart grid's stability and guarantee that it operates in a safe state. Artificial intelligence and Machine learning algorithms have proven to be effective measures to accurately predict the smart grid's stability. Despite the presence of known adversarial attacks and potential solutions, currently, there exists no standardized measure to protect smart grids against this threat, leaving them open to new adversarial attacks. In this paper, we propose GAN-GRID a novel adversarial attack targeting the stability prediction system of a smart grid tailored to real-world constraints. Our findings reveal that an adversary armed solely with the stability model's output, devoid of data or model knowledge, can craft data classified as stable with an Attack Success Rate (ASR) of 0.99. Also by manipulating authentic data and sensor values, the attacker can amplify grid issues, potentially undetected due to a compromised stability prediction system. These results underscore the imperative of fortifying smart grid security mechanisms against adversarial manipulation to uphold system stability and reliability.

We present a study of spin-unpolarized and spin-polarized two-dimensional uniform electron liquids using variational and diffusion quantum Monte Carlo (VMC and DMC) methods with Slater-Jastrow-backflow trial wave functions. Ground-state VMC and DMC energies are obtained in the density range $1 \leq r_\text{s} \leq 40$. Single-particle and many-body finite-size errors are corrected using canonical-ensemble twist-averaged boundary conditions and extrapolation of twist-averaged energies to the thermodynamic limit of infinite system size. System-size-dependent errors in Slater-Jastrow-backflow DMC energies caused by partially converged VMC energy minimization calculations are discussed. We find that, for $1 \leq r_\text{s} \leq 5$, optimizing the backflow function at each twist lowers the twist-averaged DMC energy at finite system size. However, nonsystematic system-size-dependent effects remain in the DMC energies, which can be partially removed by extrapolation from multiple finite system sizes to infinite system size. We attribute these nonsystematic effects to the close competition between fluid and defected crystal phases at different system sizes at low density. The DMC energies in the thermodynamic limit are used to parameterize a local spin density approximation correlation functional for inhomogeneous electron systems. Our zero-temperature phase diagram shows a single transition from a paramagnetic fluid to a hexagonal Wigner crystal at $r_\text{s}=35(1)$, with no region of stability for a ferromagnetic fluid.

While supermassive black holes (BHs) are widely observed in the nearby and distant universe, their origin remains debated with two viable formation scenarios with light and heavy seeds. In the light seeding model, the first BHs form from the collapse of massive stars with masses of $10-100 \ \rm{M_{\odot}}$, while the heavy seeding model posits the formation of $10^{4-5} \ \rm{M_{\odot}}$ seeds from direct collapse. The detection of BHs at redshifts $z\gtrsim10$, edging closer to their formation epoch, provides critical observational discrimination between these scenarios. Here, we focus on the JWST-detected galaxy, GHZ 9, at $z\approx10$ that is lensed by the foreground cluster, Abell 2744. Based on 2.1 Ms deep Chandra observations, we detect a candidate X-ray AGN, which is spatially coincident with the high-redshift galaxy, GHZ 9. The BH candidate is inferred to have a bolometric luminosity of $(1.0^{+0.5}_{-0.4})\times10^{46} \ \rm{erg \ s^{-1}}$, which corresponds to a BH mass of $(8.0^{+3.7}_{-3.2})\times10^7 \ \rm{M_{\odot}}$ assuming Eddington-limited accretion. This extreme mass at such an early cosmic epoch suggests the heavy seed origin for this BH candidate. Based on the Chandra and JWST discoveries of extremely high-redshift quasars, we have constructed the first simple AGN luminosity function extending to $z\approx10$. Comparison of this luminosity function with theoretical models indicates an over-abundant $z\approx10$ BH population, consistent with a higher-than-expected seed formation efficiency., Comment: 9 pages, 4 figures, accepted for publication in The Astrophysical Journal Letters

Nuclear Magnetic Resonance (NMR) spectroscopy is essential for revealing molecular structure, electronic environment, and dynamics. Accurate NMR shift prediction allows researchers to validate structures by comparing predicted and observed shifts. While Machine Learning (ML) has improved one-dimensional (1D) NMR shift prediction, predicting 2D NMR remains challenging due to limited annotated data. To address this, we introduce an unsupervised training framework for predicting cross-peaks in 2D NMR, specifically Heteronuclear Single Quantum Coherence (HSQC).Our approach pretrains an ML model on an annotated 1D dataset of 1H and 13C shifts, then finetunes it in an unsupervised manner using unlabeled HSQC data, which simultaneously generates cross-peak annotations. Our model also adjusts for solvent effects. Evaluation on 479 expert-annotated HSQC spectra demonstrates our model's superiority over traditional methods (ChemDraw and Mestrenova), achieving Mean Absolute Errors (MAEs) of 2.05 ppm and 0.165 ppm for 13C shifts and 1H shifts respectively. Our algorithmic annotations show a 95.21% concordance with experts' assignments, underscoring the approach's potential for structural elucidation in fields like organic chemistry, pharmaceuticals, and natural products.

Accurate classification of objects in 3D point clouds is a significant problem in several applications, such as autonomous navigation and augmented/virtual reality scenarios, which has become a research hot spot. In this paper, we presented a deep learning strategy for 3D object classification in augmented reality. The proposed approach is a combination of the GRU and LSTM. LSTM networks learn longer dependencies well, but due to the number of gates, it takes longer to train; on the other hand, GRU networks have a weaker performance than LSTM, but their training speed is much higher than GRU, which is The speed is due to its fewer gates. The proposed approach used the combination of speed and accuracy of these two networks. The proposed approach achieved an accuracy of 0.99 in the 4,499,0641 points dataset, which includes eight classes (unlabeled, man-made terrain, natural terrain, high vegetation, low vegetation, buildings, hardscape, scanning artifacts, cars). Meanwhile, the traditional machine learning approaches could achieve a maximum accuracy of 0.9489 in the best case. Keywords: Point Cloud Classification, Virtual Reality, Hybrid Model, GRULSTM, GRU, LSTM

We present in-situ x-ray diffraction and velocity measurements of Fe$_2$O$_3$ under laser shock compression at pressures between 38-116 GPa. None of the phases reported by static compression studies were observed. Instead, we observed an isostructural phase transition from $\alpha$-Fe$_2$O$_3$ to a new $\alpha^\prime$-Fe$_2$O$_3$ phase at a pressure of 50-62 GPa. The $\alpha^\prime$-Fe$_2$O$_3$ phase differs from $\alpha$-Fe$_2$O$_3$ by an 11% volume drop and a different unit cell compressibility. We further observed a two-wave structure in the velocity profile, which can be related to an intermediate regime where both $\alpha$ and $\alpha^\prime$ phases coexist. Density functional theory calculations with a Hubbard parameter indicate that the observed unit cell volume drop can be associated with a spin transition following a magnetic collapse., Comment: 7 pages, 5 figures

Resonant inelastic x-ray scattering (RIXS) is a widely used spectroscopic technique, providing access to the electronic structure and dynamics of atoms, molecules, and solids. However, RIXS requires a narrow bandwidth x-ray probe to achieve high spectral resolution. The challenges in delivering an energetic monochromated beam from an x-ray free electron laser (XFEL) thus limit its use in few-shot experiments, including for the study of high energy density systems. Here we demonstrate that by correlating the measurements of the self-amplified spontaneous emission (SASE) spectrum of an XFEL with the RIXS signal, using a dynamic kernel deconvolution with a neural surrogate, we can achieve electronic structure resolutions substantially higher than those normally afforded by the bandwidth of the incoming x-ray beam. We further show how this technique allows us to discriminate between the valence structures of Fe and Fe$_2$O$_3$, and provides access to temperature measurements as well as M-shell binding energies estimates in warm-dense Fe compounds.