Location Intelligence (LI), the science of transforming location-centric geospatial data into actionable knowledge, has become a cornerstone of modern spatial decision-making. The rapid evolution of Geospatial Representation Learning is fundamentally reshaping LI development through two successive technological revolutions: the deep learning breakthrough and the emerging large language model (LLM) paradigm. While deep neural networks (DNNs) have demonstrated remarkable success in automated feature extraction from structured geospatial data (e.g., satellite imagery, GPS trajectories), the recent integration of LLMs introduces transformative capabilities for cross-modal geospatial reasoning and unstructured geo-textual data processing. This survey presents a comprehensive review of geospatial representation learning across both technological eras, organizing them into a structured taxonomy based on the complete pipeline comprising: (1) data perspective, (2) methodological perspective and (3) application perspective. We also highlight current advancements, discuss existing limitations, and propose potential future research directions in the LLM era. This work offers a thorough exploration of the field and providing a roadmap for further innovation in LI. The summary of the up-to-date paper list can be found in this https URL and will undergo continuous updates.

The proliferation of online toxic speech is a pertinent problem posing threats to demographic groups. While explicit toxic speech contains offensive lexical signals, implicit one consists of coded or indirect language. Therefore, it is crucial for models not only to detect implicit toxic speech but also to explain its toxicity. This draws a unique need for unified frameworks that can effectively detect and explain implicit toxic speech. Prior works mainly formulated the task of toxic speech detection and explanation as a text generation problem. Nonetheless, models trained using this strategy can be prone to suffer from the consequent error propagation problem. Moreover, our experiments reveal that the detection results of such models are much lower than those that focus only on the detection task. To bridge these gaps, we introduce ToXCL, a unified framework for the detection and explanation of implicit toxic speech. Our model consists of three modules: a (i) Target Group Generator to generate the targeted demographic group(s) of a given post; an (ii) Encoder-Decoder Model in which the encoder focuses on detecting implicit toxic speech and is boosted by a (iii) Teacher Classifier via knowledge distillation, and the decoder generates the necessary explanation. ToXCL achieves new state-of-the-art effectiveness, and outperforms baselines significantly.

We report the first measurement of charged particle elliptic flow in Pb-Pb collisions at 2.76 TeV with the ALICE detector at the CERN Large Hadron Collider. The measurement is performed in the central pseudorapidity region (|$\eta$|<0.8) and transverse momentum range 0.2< $p_{\rm T}$< 5.0 GeV/$c$. The elliptic flow signal v$_2$, measured using the 4-particle correlation method, averaged over transverse momentum and pseudorapidity is 0.087 $\pm$ 0.002 (stat) $\pm$ 0.004 (syst) in the 40-50% centrality class. The differential elliptic flow v$_2(p_{\rm T})$ reaches a maximum of 0.2 near $p_{\rm T}$ = 3 GeV/$c$. Compared to RHIC Au-Au collisions at 200 GeV, the elliptic flow increases by about 30%. Some hydrodynamic model predictions which include viscous corrections are in agreement with the observed increase.

Recent James Webb Space Telescope (JWST) observations have revealed a population of sources with a compact morphology and a `v-shaped' continuum, namely blue at rest-frame \lambda&lt;4000A and red at longer wavelengths. The nature of these sources, called `little red dots' (LRDs), is still debated, since it is unclear if they host active galactic nuclei (AGN) and their number seems to drastically drop at z<4. We utilise the 63 $deg^2$ covered by the quick Euclid Quick Data Release (Q1) to extend the search for LRDs to brighter magnitudes and to lower z than what has been possible with JWST to have a broader view of the evolution of this peculiar galaxy population. The selection is done by fitting the available photometric data (Euclid, Spitzer/IRAC, and ground-based griz data) with two power laws, to retrieve the rest-frame optical and UV slopes consistently over a large redshift range (i.e, z<7.6). We exclude extended objects and possible line emitters, and perform a visual inspection to remove imaging artefacts. The final selection includes 3341 LRD candidates from z=0.33 to z=3.6, with 29 detected in IRAC. Their rest-frame UV luminosity function, in contrast with previous JWST studies, shows that the number density of LRD candidates increases from high-z down to z=1.5-2.5 and decreases at even lower z. Less evolution is apparent focusing on the subsample of more robust LRD candidates having IRAC detections, which is affected by low statistics and limited by the IRAC resolution. The comparison with previous quasar UV luminosity functions shows that LRDs are not the dominant AGN population at z<4. Follow-up studies of these LRD candidates are key to confirm their nature, probe their physical properties and check for their compatibility with JWST sources, since the different spatial resolution and wavelength coverage of Euclid and JWST could select different samples of compact sources.

We perform a search for light sterile neutrinos using the data from the T2K far detector at a baseline of 295 km, with an exposure of 14.7 (7.6)$\times 10^{20}$ protons on target in neutrino (antineutrino) mode. A selection of neutral current interaction samples are also used to enhance the sensitivity to sterile mixing. No evidence of sterile neutrino mixing in the 3+1 model was found from a simultaneous fit to the charged-current muon, electron and neutral current neutrino samples. We set the most stringent limit on the sterile oscillation amplitude $\sin^2\theta_{24}$ for the sterile neutrino mass splitting \Delta m^2_{41}&lt;3\times 10^{-3} eV$^2/c^4$.

Accurate 3D perception is essential for autonomous driving. Traditional methods often struggle with geometric ambiguity due to a lack of geometric prior. To address these challenges, we use omnidirectional depth estimation to introduce geometric prior. Based on the depth information, we propose a Sketch-Coloring framework OmniDepth-Occ. Additionally, our approach introduces a cylindrical voxel representation based on polar coordinate to better align with the radial nature of panoramic camera views. To address the lack of fisheye camera dataset in autonomous driving tasks, we also build a virtual scene dataset with six fisheye cameras, and the data volume has reached twice that of SemanticKITTI. Experimental results demonstrate that our Sketch-Coloring network significantly enhances 3D perception performance.

Recent advances in one-shot imitation learning have enabled robots to acquire new manipulation skills from a single human demonstration. While existing methods achieve strong performance on single-step tasks, they remain limited in their ability to handle long-horizon, multi-step tasks without additional model training or manual annotation. We propose a method that can be applied to this setting provided a single demonstration without additional model training or manual annotation. We evaluated our method on multi-step and single-step manipulation tasks where our method achieves an average success rate of 82.5% and 90%, respectively. Our method matches and exceeds the performance of the baselines in both these cases. We also compare the performance and computational efficiency of alternative pre-trained feature extractors within our framework.

The six parameters of the standard $\Lambda$CDM model have best-fit values derived from the Planck temperature power spectrum that are shifted somewhat from the best-fit values derived from WMAP data. These shifts are driven by features in the Planck temperature power spectrum at angular scales that had never before been measured to cosmic-variance level precision. We investigate these shifts to determine whether they are within the range of expectation and to understand their origin in the data. Taking our parameter set to be the optical depth of the reionized intergalactic medium $\tau$, the baryon density $\omega_{\rm b}$, the matter density $\omega_{\rm m}$, the angular size of the sound horizon $\theta_*$, the spectral index of the primordial power spectrum, $n_{\rm s}$, and $A_{\rm s}e^{-2\tau}$ (where $A_{\rm s}$ is the amplitude of the primordial power spectrum), we examine the change in best-fit values between a WMAP-like large angular-scale data set (with multipole moment \ell&lt;800 in the Planck temperature power spectrum) and an all angular-scale data set (\ell&lt;2500 Planck temperature power spectrum), each with a prior on $\tau$ of $0.07\pm0.02$. We find that the shifts, in units of the 1$\sigma$ expected dispersion for each parameter, are $\{\Delta \tau, \Delta A_{\rm s} e^{-2\tau}, \Delta n_{\rm s}, \Delta \omega_{\rm m}, \Delta \omega_{\rm b}, \Delta \theta_*\} = \{-1.7, -2.2, 1.2, -2.0, 1.1, 0.9\}$, with a$\chi^2$ value of 8.0. We find that this $\chi^2$ value is exceeded in 15% of our simulated data sets, and that a parameter deviates by more than 2.2$\sigma$ in 9% of simulated data sets, meaning that the shifts are not unusually large. Comparing \ell&lt;800 instead to \ell&gt;800, or splitting at a different multipole, yields similar results. We examine the \ell&lt;800 model residuals in the \ell&gt;800 power spectrum data and find that the features there... [abridged]

Language Models (LMs) excel in natural language processing tasks for English but show reduced performance in most other languages. This problem is commonly tackled by continually pre-training and fine-tuning these models for said languages. A significant issue in this process is the limited vocabulary coverage in the original model's tokenizer, leading to inadequate representation of new languages and necessitating an expansion of the tokenizer. The initialization of the embeddings corresponding to new vocabulary items presents a further challenge. Current strategies require cross-lingual embeddings and lack a solid theoretical foundation as well as comparisons with strong baselines. In this paper, we first establish theoretically that initializing within the convex hull of existing embeddings is a good initialization, followed by a novel but simple approach, Constrained Word2Vec (CW2V), which does not require cross-lingual embeddings. Our study evaluates different initialization methods for expanding RoBERTa and LLaMA 2 across four languages and five tasks. The results show that CW2V performs equally well or even better than more advanced techniques. Additionally, simpler approaches like multivariate initialization perform on par with these advanced methods indicating that efficient large-scale multilingual continued pretraining can be achieved even with simpler initialization methods. We release our code publicly (this https URL).

Clinical evidence has shown that rib-suppressed chest X-rays (CXRs) can improve the reliability of pulmonary disease diagnosis. However, previous approaches on generating rib-suppressed CXR face challenges in preserving details and eliminating rib residues. We hereby propose a GAN-based disentanglement learning framework called Rib Suppression GAN, or RSGAN, to perform rib suppression by utilizing the anatomical knowledge embedded in unpaired computed tomography (CT) images. In this approach, we employ a residual map to characterize the intensity difference between CXR and the corresponding rib-suppressed result. To predict the residual map in CXR domain, we disentangle the image into structure- and contrast-specific features and transfer the rib structural priors from digitally reconstructed radiographs (DRRs) computed by CT. Furthermore, we employ additional adaptive loss to suppress rib residue and preserve more details. We conduct extensive experiments based on 1,673 CT volumes, and four benchmarking CXR datasets, totaling over 120K images, to demonstrate that (i) our proposed RSGAN achieves superior image quality compared to the state-of-the-art rib suppression methods; (ii) combining CXR with our rib-suppressed result leads to better performance in lung disease classification and tuberculosis area detection.

The IceCube Collaboration has previously discovered a high-energy astrophysical neutrino flux using neutrino events with interaction vertices contained within the instrumented volume of the IceCube detector. We present a complementary measurement using charged current muon neutrino events where the interaction vertex can be outside this volume. As a consequence of the large muon range the effective area is significantly larger but the field of view is restricted to the Northern Hemisphere. IceCube data from 2009 through 2015 have been analyzed using a likelihood approach based on the reconstructed muon energy and zenith angle. At the highest neutrino energies between 191 TeV and 8.3 PeV a significant astrophysical contribution is observed, excluding a purely atmospheric origin of these events at $5.6\,\sigma$ significance. The data are well described by an isotropic, unbroken power law flux with a normalization at 100 TeV neutrino energy of $\left(0.90^{+0.30}_{-0.27}\right)\times10^{-18}\,\mathrm{GeV^{-1}\,cm^{-2}\,s^{-1}\,sr^{-1}}$ and a hard spectral index of $\gamma=2.13\pm0.13$. The observed spectrum is harder in comparison to previous IceCube analyses with lower energy thresholds which may indicate a break in the astrophysical neutrino spectrum of unknown origin. The highest energy event observed has a reconstructed muon energy of $(4.5\pm1.2)\,\mathrm{PeV}$ which implies a probability of less than 0.005% for this event to be of atmospheric origin. Analyzing the arrival directions of all events with reconstructed muon energies above 200 TeV no correlation with known $\gamma$-ray sources was found. Using the high statistics of atmospheric neutrinos we report the currently best constraints on a prompt atmospheric muon neutrino flux originating from charmed meson decays which is below $1.06$ in units of the flux normalization of the model in Enberg et al. (2008).

Let $\Gamma\in \mathrm{SL}_2(\mathbb{C})$ be a finite subgroup. We introduce a class of projective noncommutative surfaces $\mathbb{P}^2_I$, indexed by a set of irreducible $\Gamma$-representations. Extending the action of $\Gamma$ from $\mathbb{C}^2$ to $\mathbb{P}^2$, we show that these surfaces generalise both $[\mathbb{P}^2/\Gamma]$ and $\mathbb{P}^2/\Gamma$. We prove that isomorphism classes of framed torsion-free sheaves on any $\mathbb{P}^2_I$ carry a canonical bijection to the closed points of appropriate Nakajima quiver varieties. In particular, we provide geometric interpretations for a class of Nakajima quiver varieties using noncommutative geometry. Our results partially generalise several previous results on such quiver varieties.

Applying deep neural networks to 3D point cloud processing has attracted increasing attention due to its advanced performance in many areas, such as AR/VR, autonomous driving, and robotics. However, as neural network models and 3D point clouds expand in size, it becomes a crucial challenge to reduce the computational and memory overhead to meet latency and energy constraints in real-world applications. Although existing approaches have proposed to reduce both computational cost and memory footprint, most of them only address the spatial redundancy in inputs, i.e. removing the redundancy of background points in 3D data. In this paper, we propose a novel post-training weight pruning scheme for 3D object detection that is (1) orthogonal to all existing point cloud sparsifying methods, which determines redundant parameters in the pretrained model that lead to minimal distortion in both locality and confidence (detection distortion); and (2) a universal plug-and-play pruning framework that works with arbitrary 3D detection model. This framework aims to minimize detection distortion of network output to maximally maintain detection precision, by identifying layer-wise sparsity based on second-order Taylor approximation of the distortion. Albeit utilizing second-order information, we introduced a lightweight scheme to efficiently acquire Hessian information, and subsequently perform dynamic programming to solve the layer-wise sparsity. Extensive experiments on KITTI, Nuscenes and ONCE datasets demonstrate that our approach is able to maintain and even boost the detection precision on pruned model under noticeable computation reduction (FLOPs). Noticeably, we achieve over 3.89x, 3.72x FLOPs reduction on CenterPoint and PVRCNN model, respectively, without mAP decrease, significantly improving the state-of-the-art.

We propose to tackle the problem of multiview 2D/3D rigid registration for intervention via a Point-Of-Interest Network for Tracking and Triangulation ($\text{POINT}^2$). $\text{POINT}^2$ learns to establish 2D point-to-point correspondences between the pre- and intra-intervention images by tracking a set of random POIs. The 3D pose of the pre-intervention volume is then estimated through a triangulation layer. In $\text{POINT}^2$, the unified framework of the POI tracker and the triangulation layer enables learning informative 2D features and estimating 3D pose jointly. In contrast to existing approaches, $\text{POINT}^2$ only requires a single forward-pass to achieve a reliable 2D/3D registration. As the POI tracker is shift-invariant, $\text{POINT}^2$ is more robust to the initial pose of the 3D pre-intervention image. Extensive experiments on a large-scale clinical cone-beam CT (CBCT) dataset show that the proposed $\text{POINT}^2$ method outperforms the existing learning-based method in terms of accuracy, robustness and running time. Furthermore, when used as an initial pose estimator, our method also improves the robustness and speed of the state-of-the-art optimization-based approaches by ten folds.

Computed tomography (CT) reconstruction from X-ray projections acquired within a limited angle range is challenging, especially when the angle range is extremely small. Both analytical and iterative models need more projections for effective modeling. Deep learning methods have gained prevalence due to their excellent reconstruction performances, but such success is mainly limited within the same dataset and does not generalize across datasets with different distributions. Hereby we propose ExtraPolationNetwork for limited-angle CT reconstruction via the introduction of a sinogram extrapolation module, which is theoretically justified. The module complements extra sinogram information and boots model generalizability. Extensive experimental results show that our reconstruction model achieves state-of-the-art performance on NIH-AAPM dataset, similar to existing approaches. More importantly, we show that using such a sinogram extrapolation module significantly improves the generalization capability of the model on unseen datasets (e.g., COVID-19 and LIDC datasets) when compared to existing approaches.

The cross sections of $e^{+}e^{-}\to\pi^{+}\pi^{-}h_c$ at center-of-mass energies from 3.896 to 4.600 GeV are measured using data samples collected with the BESIII detector operating at the Beijing Electron Positron Collider. The cross sections are found to be of the same order of magnitude as those of $e^{+}e^{-}\to\pi^{+}\pi^{-}J/\psi$ and $e^{+}e^{-}\to\pi^{+}\pi^{-}\psi(2S)$, but the line shape is inconsistent with the $Y$ states observed in the latter two modes. Two structures are observed in the $e^{+}e^{-}\to\pi^{+}\pi^{-}h_c$ cross sections around 4.22 and 4.39 GeV/$c^{2}$, which we call $Y(4220)$ and $Y(4390)$, respectively. A fit with a coherent sum of two Breit-Wigner functions results in a mass of $(4218.4^{+5.5}_{-4.5}\pm0.9)$ MeV/$c^{2}$ and a width of $(66.0^{+12.3}_{-8.3}\pm0.4)$ MeV for the $Y(4220)$, and a mass of $(4391.6^{+6.3}_{-6.8}\pm1.0)$ MeV/$c^{2}$ and a width of $(139.5^{+16.2}_{-20.6}\pm0.6)$ MeV for the $Y(4390)$, where the first uncertainties are statistical and the second ones systematic. The statistical significance of $Y(4220)$ and $Y(4390)$ is 10$\sigma$ over one structure assumption.

This document presents the concept and physics case for a magnetized gaseous argon-based detector system (ND-GAr) for the Deep Underground Neutrino Experiment (DUNE) Near Detector. This detector system is required in order for DUNE to reach its full physics potential in the measurement of CP violation and in delivering precision measurements of oscillation parameters. In addition to its critical role in the long-baseline oscillation program, ND-GAr will extend the overall physics program of DUNE. The LBNF high-intensity proton beam will provide a large flux of neutrinos that is sampled by ND-GAr, enabling DUNE to discover new particles and search for new interactions and symmetries beyond those predicted in the Standard Model.

DarkSide-20k (DS-20k) is a dark matter detection experiment under construction at the Laboratori Nazionali del Gran Sasso (LNGS) in Italy. It utilises ~100 t of low radioactivity argon from an underground source (UAr) in its inner detector, with half serving as target in a dual-phase time projection chamber (TPC). The UAr cryogenics system must maintain stable thermodynamic conditions throughout the experiment's lifetime of over 10 years. Continuous removal of impurities and radon from the UAr is essential for maximising signal yield and mitigating background. We are developing an efficient and powerful cryogenics system with a gas purification loop with a target circulation rate of 1000 slpm. Central to its design is a condenser operated with liquid nitrogen which is paired with a gas heat exchanger cascade, delivering a combined cooling power of more than 8 kW. Here we present the design choices in view of the DS-20k requirements, in particular the condenser's working principle and the cooling control, and we show test results obtained with a dedicated benchmarking platform at CERN and LNGS. We find that the thermal efficiency of the recirculation loop, defined in terms of nitrogen consumption per argon flow rate, is 95 % and the pressure in the test cryostat can be maintained within $\pm$(0.1-0.2) mbar. We further detail a 5-day cool-down procedure of the test cryostat, maintaining a cooling rate typically within -2 K/h, as required for the DS-20k inner detector. Additionally, we assess the circuit's flow resistance, and the heat transfer capabilities of two heat exchanger geometries for argon phase change, used to provide gas for recirculation. We conclude by discussing how our findings influence the finalisation of the system design, including necessary modifications to meet requirements and ongoing testing activities.

The Agent Based Model community has a rich and diverse ecosystem of libraries, platforms, and applications to help modelers develop rigorous simulations. Despite this robust and diverse ecosystem, the complexity of life from microbial communities to the global ecosystem still presents substantial challenges in making reusable code that can optimize the ability of the knowledge-sharing and reproducibility. This research seeks to provide new tools to mitigate some of these challenges by offering a vision of a more holistic ecosystem that takes researchers and practitioners from the data collection through validation, with transparent, accessible, and extensible subcomponents. This proposed approach is demonstrated through two data pipelines (crop yield and synthetic population) that take users from data download through the cleaning and processing until users of have data that can be integrated into an ABM. These pipelines are built to be transparent: by walking users step by step through the process, accessible: by being skill scalable so users can leverage them without code or with code, and extensible by being freely available on the coding sharing repository GitHub to facilitate community development. Reusing code that simulates complex phenomena is a significant challenge but one that must be consistently addressed to help the community move forward. This research seeks to aid that progress by offering potential new tools extended from the already robust ecosystem to help the community collaborate more effectively internally and across disciplines.