Search

We present Simplified Text-Attributed Graph Embeddings (STAGE), a straightforward yet effective method for enhancing node features in Graph Neural Network (GNN) models that encode Text-Attributed Graphs (TAGs). Our approach leverages Large-Language Models (LLMs) to generate embeddings for textual attributes. STAGE achieves competitive results on various node classification benchmarks while also maintaining a simplicity in implementation relative to current state-of-the-art (SoTA) techniques. We show that utilizing pre-trained LLMs as embedding generators provides robust features for ensemble GNN training, enabling pipelines that are simpler than current SoTA approaches which require multiple expensive training and prompting stages. We also implement diffusion-pattern GNNs in an effort to make this pipeline scalable to graphs beyond academic benchmarks.

By prior work, we have many results related to distributed graph algorithms for problems that can be defined with local constraints; the formal framework used in prior work is locally checkable labeling problems (LCLs), introduced by Naor and Stockmeyer in the 1990s. It is known, for example, that if we have a deterministic algorithm that solves an LCL in $o(\log n)$ rounds, we can speed it up to $O(\log^*n)$ rounds, and if we have a randomized $O(\log^*n)$ rounds algorithm, we can derandomize it for free. It is also known that randomness helps with some LCL problems: there are LCL problems with randomized complexity $\Theta(\log\log n)$ and deterministic complexity $\Theta(\log n)$. However, so far there have not been any LCL problems in which the use of shared randomness has been necessary; in all prior algorithms it has been enough that the nodes have access to their own private sources of randomness. Could it be the case that shared randomness never helps with LCLs? Could we have a general technique that takes any distributed graph algorithm for any LCL that uses shared randomness, and turns it into an equally fast algorithm where private randomness is enough? In this work we show that the answer is no. We present an LCL problem $\Pi$ such that the round complexity of $\Pi$ is $\Omega(\sqrt n)$ in the usual randomized \local model with private randomness, but if the nodes have access to a source of shared randomness, then the complexity drops to $O(\log n)$. As corollaries, we also resolve several other open questions related to the landscape of distributed computing in the context of LCL problems. In particular, problem $\Pi$ demonstrates that distributed quantum algorithms for LCL problems strictly benefit from a shared quantum state. Problem $\Pi$ also gives a separation between finitely dependent distributions and non-signaling distributions.

We numerically study a double-slit interferometer that detects the modification of local phase and amplitude signal due to a refractive index contrast in a nanoplasmonic sensor after the plasmonic wave has been coupled to the farfield. Specifically, the sensor consists of two elongated zero-mode waveguides (ZMW) which function to launch surface plasmons as well as transmit light directly into the farfield. The surface plasmon waves are coupled to the farfield by surface gratings which cover one quarter of a bull-eye antenna, with different transmission directions. In the farfield, the coupled plasmon waves interfere with the directly transmitted reference wave, which reveals a combination of the phase and amplitude modification by the asymmetry of refractive index in the lumen of the ZMW. We report a drastic increase in sensitivity over interferometers that only utilize the directly transmitted fraction., Comment: 18 pages, 12 figures

This paper is dedicated to the construction of multidimensional spherical monogenics. Firstly, we investigate the construction of monogenic functions in dimension $3$ by applying the Dirac operator to the orthonormal bases of spherical harmonics, resulting in orthogonal spherical monogenics. Additionally, we employ the reproducing kernel for monogenic functions and a specialized optimization method to derive various types of $3$-dimensional spherical harmonics and spherical monogenics.

This study addresses the challenge of integrating social norms into robot navigation, which is essential for ensuring that robots operate safely and efficiently in human-centric environments. Social norms, often unspoken and implicitly understood among people, are difficult to explicitly define and implement in robotic systems. To overcome this, we derive these norms from real human trajectory data, utilizing the comprehensive ATC dataset to identify the minimum social zones humans and robots must respect. These zones are integrated into the robot' navigation system by applying barrier functions, ensuring the robot consistently remains within the designated safety set. Simulation results demonstrate that our system effectively mimics human-like navigation strategies, such as passing on the right side and adjusting speed or pausing in constrained spaces. The proposed framework is versatile, easily comprehensible, and tunable, demonstrating the potential to advance the development of robots designed to navigate effectively in human-centric environments.

Deployment of Large Language Models (LLMs) has major computational costs, due to their rapidly expanding size. Compression of LLMs reduces the memory footprint, latency, and energy required for their inference. Post-training Quantization (PTQ) techniques have been developed to compress LLMs while avoiding expensive re-training. Most PTQ approaches formulate the quantization error based on a layer-wise $\ell_2$ loss, ignoring the model output. Then, each layer is calibrated using its layer-wise Hessian to update the weights towards minimizing the $\ell_2$ quantization error. The Hessian is also used for detecting the most salient weights to quantization. Such PTQ approaches are prone to accuracy drop in low-precision quantization. We propose Output-adaptive Calibration (OAC) to incorporate the model output in the calibration process. We formulate the quantization error based on the distortion of the output cross-entropy loss. OAC approximates the output-adaptive Hessian for each layer under reasonable assumptions to reduce the computational complexity. The output-adaptive Hessians are used to update the weight matrices and detect the salient weights towards maintaining the model output. Our proposed method outperforms the state-of-the-art baselines such as SpQR and BiLLM, especially, at extreme low-precision (2-bit and binary) quantization., Comment: 20 pages, 4 figures

The current standard model of cosmology successfully describes a variety of measurements, but the nature of its main ingredients, dark matter and dark energy, remains unknown. Euclid is a medium-class mission in the Cosmic Vision 2015-2025 programme of the European Space Agency (ESA) that will provide high-resolution optical imaging, as well as near-infrared imaging and spectroscopy, over about 14,000 deg^2 of extragalactic sky. In addition to accurate weak lensing and clustering measurements that probe structure formation over half of the age of the Universe, its primary probes for cosmology, these exquisite data will enable a wide range of science. This paper provides a high-level overview of the mission, summarising the survey characteristics, the various data-processing steps, and data products. We also highlight the main science objectives and expected performance., Comment: Paper submitted as part of the A&A special issue`Euclid on Sky'

The ever-growing computational complexity of Large Language Models (LLMs) necessitates efficient deployment strategies. The current state-of-the-art approaches for Post-training Quantization (PTQ) often require calibration to achieve the desired accuracy. This paper presents AdpQ, a novel zero-shot adaptive PTQ method for LLMs that achieves the state-of-the-art performance in low-precision quantization (e.g. 3-bit) without requiring any calibration data. Inspired by Adaptive LASSO regression model, our proposed approach tackles the challenge of outlier activations by separating salient weights using an adaptive soft-thresholding method. Guided by Adaptive LASSO, this method ensures that the quantized weights distribution closely follows the originally trained weights and eliminates the need for calibration data entirely, setting our method apart from popular approaches such as SpQR and AWQ. Furthermore, our method offers an additional benefit in terms of privacy preservation by eliminating any calibration or training data. We also delve deeper into the information-theoretic underpinnings of the proposed method. We demonstrate that it leverages the Adaptive LASSO to minimize the Kullback-Leibler divergence between the quantized weights and the originally trained weights. This minimization ensures the quantized model retains the Shannon information content of the original model to a great extent, guaranteeing efficient deployment without sacrificing accuracy or information. Our results achieve the same accuracy as the existing methods on various LLM benchmarks while the quantization time is reduced by at least 10x, solidifying our contribution to efficient and privacy-preserving LLM deployment.

This research reports VascularPilot3D, the first 3D fully autonomous endovascular robot navigation system. As an exploration toward autonomous guidewire navigation, VascularPilot3D is developed as a complete navigation system based on intra-operative imaging systems (fluoroscopic X-ray in this study) and typical endovascular robots. VascularPilot3D adopts previously researched fast 3D-2D vessel registration algorithms and guidewire segmentation methods as its perception modules. We additionally propose three modules: a topology-constrained 2D-3D instrument end-point lifting method, a tree-based fast path planning algorithm, and a prior-free endovascular navigation strategy. VascularPilot3D is compatible with most mainstream endovascular robots. Ex-vivo experiments validate that VascularPilot3D achieves 100% success rate among 25 trials. It reduces the human surgeon's overall control loops by 18.38%. VascularPilot3D is promising for general clinical autonomous endovascular navigations., Comment: Submitted to MICCAI2024

Transferring the weights of a pre-trained model to assist another task has become a crucial part of modern deep learning, particularly in data-scarce scenarios. Pre-training refers to the initial step of training models outside the current task of interest, typically on another dataset. It can be done via supervised models using human-annotated datasets or self-supervised models trained on unlabeled datasets. In both cases, many pre-trained models are available to fine-tune for the task of interest. Interestingly, research has shown that pre-trained models from ImageNet can be helpful for audio tasks despite being trained on image datasets. Hence, it's unclear whether in-domain models would be advantageous compared to competent out-domain models, such as convolutional neural networks from ImageNet. Our experiments will demonstrate the usefulness of in-domain models and datasets for bird species recognition by leveraging VICReg, a recent and powerful self-supervised method.

Emission of terahertz waves in free electron lasers (FELs) with both planar (linear) and helical undulators has been explored. The analysis employs Lienard-Wiechert fields to characterize FEL radiation within the THz region. Specifically, the study delves into the analytical examination of radiation pattern of THz waves, which provides insight into the angular distribution of radiation energy. The variations of radiation pattern for various parameters such as different harmonics, Lorentz factor, magnetic parameter, and total number of periods of planar and helical undulators were investigated. The Fourier transform of free electron laser electric field is showcased across four distinct sets of FEL parameters. The interplay between electron beam and undulators parameters sheds light on how these factors influence the directionality of THz waves. Simulation results indicate that planar undulators often produce radiation with noticeable side lobes, especially at higher harmonics, while helical undulators typically exhibit reduced side lobes, resulting in a more focused main radiation peak. However, the radiation pattern of THz waves tends to concentrate predominantly in the forward direction and the polarization of the emitted radiation is determined by the properties of the undulator and electron beam, rather than the specific undulator geometry. The study underscores that an optimal selection of undulator and beam parameters can effectively maximize the angular distribution of emitted radiation.

This study explores the use of Large Language Models (LLMs) for automatic evaluation of knowledge graph (KG) completion models. Historically, validating information in KGs has been a challenging task, requiring large-scale human annotation at prohibitive cost. With the emergence of general-purpose generative AI and LLMs, it is now plausible that human-in-the-loop validation could be replaced by a generative agent. We introduce a framework for consistency and validation when using generative models to validate knowledge graphs. Our framework is based upon recent open-source developments for structural and semantic validation of LLM outputs, and upon flexible approaches to fact checking and verification, supported by the capacity to reference external knowledge sources of any kind. The design is easy to adapt and extend, and can be used to verify any kind of graph-structured data through a combination of model-intrinsic knowledge, user-supplied context, and agents capable of external knowledge retrieval., Comment: Text2KG 2024, ESWC 2024

We present subwavelength resonant lattices fashioned as nano- and microstructured films as a basis for a host of device concepts. Whereas the canonical physical properties are fully embodied in a one-dimensional periodic lattice, the final device constructs are often patterned in two-dimensionally-modulated films in which case we may refer to them as photonic crystal slabs, metamaterials, or metasurfaces. These surfaces can support lateral modes and localized field signatures with propagative and evanescent diffraction channels critically controlling the response. The governing principle of guided-mode, or lattice, resonance enables diverse spectral expressions such that a single-layer component can behave as a sensor, reflector, filter, or polarizer. This structural sparsity contrasts strongly with the venerable field of multi-layer thin-film optics that is basis for most optical components on the market today. The lattice resonance effect can be exploited in all major spectral regions with appropriate low-loss materials and fabrication resources. In this paper, we highlight resonant device technology and present our work on design, fabrication, and characterization of optical elements operating in the near-IR, mid-IR, and long-wave IR spectral regions. Examples of fabricated and tested devices include biological sensors, high-contrast-ratio polarizers, narrow-band notch filters, and wideband high reflectors., Comment: 11 pages, 10 figures

We revisit the classic broadcast problem, wherein we have $k$ messages, each composed of $O(\log{n})$ bits, distributed arbitrarily across a network. The objective is to broadcast these messages to all nodes in the network. In the distributed CONGEST model, a textbook algorithm solves this problem in $O(D+k)$ rounds, where $D$ is the diameter of the graph. While the $O(D)$ term in the round complexity is unavoidable$\unicode{x2014}$given that $\Omega(D)$ rounds are necessary to solve broadcast in any graph$\unicode{x2014}$it remains unclear whether the $O(k)$ term is needed in all graphs. In cases where the minimum cut size is one, simply transmitting messages from one side of the cut to the other would require $\Omega(k)$ rounds. However, if the size of the minimum cut is larger, it may be possible to develop faster algorithms. This motivates the exploration of the broadcast problem in networks with high edge connectivity. In this work, we present a simple randomized distributed algorithm for performing $k$-message broadcast in $O(((n+k)/\lambda)\log n)$ rounds in any $n$-node simple graph with edge connectivity $\lambda$. When $k = \Omega(n)$, our algorithm is universally optimal, up to an $O(\log n)$ factor, as its complexity nearly matches an information-theoretic $\Omega(k/\lambda)$ lower bound that applies to all graphs, even when the network topology is known to the algorithm. The setting $k = \Omega(n)$ is particularly interesting because several fundamental problems can be reduced to broadcasting $\Omega(n)$ messages. Our broadcast algorithm finds several applications in distributed computing, enabling $O(1)$-approximation for all distances and $(1+\epsilon)$-approximation for all cut sizes in $\tilde{O}(n/\lambda)$ rounds.

A second-order Newton-based extremum seeking (SONES) algorithm is presented to estimate directional inflection points for multivariable static maps. The design extends the first-order Newton-based extremum seeking algorithm that drives the system toward its peak point. This work provides perturbation matrices to estimate the second- and third-order derivatives necessary for implementation of the SONES. A set of conditions are provided for the probing frequencies that ensure accurate estimation of the derivatives. A differential Riccati filter is used to calculate the inverse of the third-order derivative. The local stability of the new algorithm is proven for general multivariable static maps using averaging analysis. The proposed algorithm ensures uniform convergence toward directional inflection point without requiring information about the curvature of the map and its gradient. Simulation results show the effectiveness of the proposed algorithm.

Self-focusing emerges as a nonlinear optical phenomenon resulting from an intense laser field and plasma interaction. This study investigates the self-focusing behavior of Gaussian laser beams within magnetized plasma environments utilizing a novel approach, source-dependent expansion. By employing source-dependent expansion, we explore the intricate dynamics of laser beam propagation, considering the influence of plasma density and external magnetic fields. The interplay between the beam's Gaussian profile and the self-focusing mechanism through rigorous mathematical analysis and numerical simulations, particularly in the presence of plasma-induced nonlinearities are elucidated here. Our findings reveal crucial insights into the evolution of laser beams under diverse parameters, the ponderomotive force, relativistic factors, plasma frequency, polarization states, external magnetic field, wavelength, and laser intensity. This research not only contributes to advancing our fundamental understanding of laser-plasma interactions but also holds promise for optimizing laser-driven applications.

The exploration of Terahertz (THz) waves has captivated researchers across diverse scientific disciplines such as physics, spectroscopy, chemistry, biology, and engineering, driven by the myriad applications these waves offer. Within this expansive landscape, the development of efficient and reliable THz sources stands as a paramount objective. In the pursuit of this goal, a multitude of approaches have been undertaken, with a notable contender emerging in the form of laser-induced plasma. Harnessing the advancements in ultrafast pulses, laser-induced plasma has proven to be a promising tool for generating THz waves. Its appeal lies in the robust attributes of a high power threshold, intense THz signal, and an broadband THz spectrum. This paper delves into a comprehensive review of the physics and progress underlying THz generation from laser-induced plasmas, exploring scenarios where plasmas are induced in gases, liquids, and solids. The interactions between lasers and plasmas involve complex physical processes, resulting in a variety of laser plasma scenarios for THz generation. In this review, the focus is specifically placed on classifying THz generation based on different physical mechanisms and also examines the characteristics of the emitted THz waves. By categorizing the processes, a deeper understanding of the underlying principles can be attained.

Pulse shaping provides a significant level of control and precision when optimizing laser-plasma interactions. Pulse shaping enables precise control and manipulation, resulting in enhanced energy deposition, optimized particle acceleration, controlled polarization, and exploitation of resonant effects. The present study investigates the interaction of structured light with magnetized plasma, considering various spatial profiles and polarization states. This phenomenon involves modification of the temporal and spatial characteristics of the laser pulse due to the presence of the magnetized plasma. The discussion reveals how the electric field and electron velocity evolve within the plasma both spatially and temporally. Factors such as absorption, dispersion, collisions, and scattering are taken into account to understand how they influence the evolution of the pulse. The effects of electron density, external magnetic fields, relativistic velocities, and polarization states on pulse compression are examined. The spatial laser profile impact on pulse-shaping and plasma channel formation is also discussed. This exploration sheds light on the intricate interplays and potential pulse-shaping applications in laser-plasma interactions., Comment: 8 figures

This paper investigates the robot state estimation problem within a non-inertial environment. The proposed state estimation approach relaxes the common assumption of static ground in the system modeling. The process and measurement models explicitly treat the movement of the non-inertial environments without requiring knowledge of its motion in the inertial frame or relying on GPS or sensing environmental landmarks. Further, the proposed state estimator is formulated as an invariant extended Kalman filter (InEKF) with the deterministic part of its process model obeying the group-affine property, leading to log-linear error dynamics. The observability analysis of the filter confirms that the robot's pose (i.e., position and orientation) and velocity relative to the non-inertial environment are observable. Hardware experiments on a humanoid robot moving on a rotating and translating treadmill demonstrate the high convergence rate and accuracy of the proposed InEKF even under significant treadmill pitch sway, as well as large estimation errors.

Medical image classification requires labeled, task-specific datasets which are used to train deep learning networks de novo, or to fine-tune foundation models. However, this process is computationally and technically demanding. In language processing, in-context learning provides an alternative, where models learn from within prompts, bypassing the need for parameter updates. Yet, in-context learning remains underexplored in medical image analysis. Here, we systematically evaluate the model Generative Pretrained Transformer 4 with Vision capabilities (GPT-4V) on cancer image processing with in-context learning on three cancer histopathology tasks of high importance: Classification of tissue subtypes in colorectal cancer, colon polyp subtyping and breast tumor detection in lymph node sections. Our results show that in-context learning is sufficient to match or even outperform specialized neural networks trained for particular tasks, while only requiring a minimal number of samples. In summary, this study demonstrates that large vision language models trained on non-domain specific data can be applied out-of-the box to solve medical image-processing tasks in histopathology. This democratizes access of generalist AI models to medical experts without technical background especially for areas where annotated data is scarce., Comment: 40 pages, 5 figures

This paper develops a new filtering approach for state estimation in polynomial systems corrupted by arbitrary noise, which commonly arise in robotics. We first consider a batch setup where we perform state estimation using all data collected from the initial to the current time. We formulate the batch state estimation problem as a Polynomial Optimization Problem (POP) and relax the assumption of Gaussian noise by specifying a finite number of moments of the noise. We solve the resulting POP using a moment relaxation and prove that under suitable conditions on the rank of the relaxation, (i) we can extract a provably optimal estimate from the moment relaxation, and (ii) we can obtain a belief representation from the dual (sum-of-squares) relaxation. We then turn our attention to the filtering setup and apply similar insights to develop a GMKF for recursive state estimation in polynomial systems with arbitrary noise. The GMKF formulates the prediction and update steps as POPs and solves them using moment relaxations, carrying over a possibly non-Gaussian belief. In the linear-Gaussian case, GMKF reduces to the standard Kalman Filter. We demonstrate that GMKF performs well under highly non-Gaussian noise and outperforms common alternatives, including the Extended and Unscented Kalman Filter, and their variants on matrix Lie group.

This work reports a novel Bundle Adjustment (BA) formulation using a Reproducing Kernel Hilbert Space (RKHS) representation called RKHS-BA. The proposed formulation is correspondence-free, enables the BA to use RGB-D/LiDAR and semantic labels in the optimization directly, and provides a generalization for the photometric loss function commonly used in direct methods. RKHS-BA can incorporate appearance and semantic labels within a continuous spatial-semantic functional representation that does not require optimization via image pyramids. We demonstrate its applications in sliding-window odometry and global LiDAR mapping, which show highly robust performance in extremely challenging scenes and the best trade-off of generalization and accuracy., Comment: 16 pages, 12 figures, technical report under review

In this paper, we present an exploration of LLMs' abilities to problem solve with physical reasoning in situated environments. We construct a simple simulated environment and demonstrate examples of where, in a zero-shot setting, both text and multimodal LLMs display atomic world knowledge about various objects but fail to compose this knowledge in correct solutions for an object manipulation and placement task. We also use BLIP, a vision-language model trained with more sophisticated cross-modal attention, to identify cases relevant to object physical properties that that model fails to ground. Finally, we present a procedure for discovering the relevant properties of objects in the environment and propose a method to distill this knowledge back into the LLM., Comment: 10 pages, 10 figures, Proceedings of AAAI Spring Symposium: Empowering Machine Learning and Large Language Models with Domain and Commonsense Knowledge (MAKE). AAAI (2024)

The number of published Machine Learning Interatomic Potentials (MLIPs) has increased significantly in recent years. These new data-driven potential energy approximations often lack the physics-based foundations that inform many traditionally-developed interatomic potentials and hence require robust validation methods for their applicability, accuracy, computational efficiency, and transferability to the intended applications. This work presents a sequential, three-stage workflow for MLIP validation: (i) preliminary validation, (ii) static property prediction, and (iii) dynamic property prediction. This material-agnostic procedure is demonstrated in a tutorial approach for the development of a robust MLIP for boron carbide (B4C), a widely employed, structurally complex ceramic that undergoes a deleterious deformation mechanism called "amorphization" under high-pressure loading. It is shown that the resulting B4C MLIP offers a more accurate prediction of properties compared to the available empirical potential.

Background: The development of smart nanocarriers that enable controlled drug release in response to internal and external triggers is an emerging approach for targeted therapy. This study focused on designing pH-sensitive, ultrasound-responsive gelatin/perfluorohexane (PFH) nanodroplets loaded with berberine chloride as a model drug. Results:The nanodroplets were prepared using an emulsion technique and optimized by varying process parameters like homogenization rate, polymer concentration, surfactant, drug, and perfluorocarbon content. The optimal formulation yielded nanodroplets with a particle size of 281.7 nm, a drug encapsulation efficiency of 66.8+- 1.7%, and a passive drug release of 15.4+- 0.2% within 24 hours. Characterization confirmed successful encapsulation and pH-responsive behavior. Ultrasound stimulation significantly enhanced drug release, with 150 kHz being more effective than 1 MHz in triggering acoustic droplet vaporization while minimizing heat generation. After 10 minutes of radiation, the optimal formulation showed 89.4% cumulative drug release. The nanodroplets displayed stability over one month at 4{\deg}C. Conclusions: Overall, the dual-triggered nanodroplets demonstrate excellent potential for controlled delivery and targeted release of berberine chloride., Comment: 15 pages , published in Bulletin of the National Research Centre journal

A backward stable numerical calculation of a function with condition number $\kappa$ will have a relative accuracy of $\kappa\epsilon_{\text{machine}}$. Standard formulations and software implementations of finite-strain elastic materials models make use of the deformation gradient $\boldsymbol F = I + \partial \boldsymbol u/\partial \boldsymbol X$ and Cauchy-Green tensors. These formulations are not numerically stable, leading to loss of several digits of accuracy when used in the small strain regime, and often precluding the use of single precision floating point arithmetic. We trace the source of this instability to specific points of numerical cancellation, interpretable as ill-conditioned steps. We show how to compute various strain measures in a stable way and how to transform common constitutive models to their stable representations, formulated in either initial or current configuration. The stable formulations all provide accuracy of order $\epsilon_{\text{machine}}$. In many cases, the stable formulations have elegant representations in terms of appropriate strain measures and offer geometric intuition that is lacking in their standard representation. We show that algorithmic differentiation can stably compute stresses so long as the strain energy is expressed stably, and give principles for stable computation that can be applied to inelastic materials.

Purpose: Common dense stereo Simultaneous Localization and Mapping (SLAM) approaches in Minimally Invasive Surgery (MIS) require high-end parallel computational resources for real-time implementation. Yet, it is not always feasible since the computational resources should be allocated to other tasks like segmentation, detection, and tracking. To solve the problem of limited parallel computational power, this research aims at a lightweight dense stereo SLAM system that works on a single-core CPU and achieves real-time performance (more than 30 Hz in typical scenarios). Methods: A new dense stereo mapping module is integrated with the ORB-SLAM2 system and named BDIS-SLAM. Our new dense stereo mapping module includes stereo matching and 3D dense depth mosaic methods. Stereo matching is achieved with the recently proposed CPU-level real-time matching algorithm Bayesian Dense Inverse Searching (BDIS). A BDIS-based shape recovery and a depth mosaic strategy are integrated as a new thread and coupled with the backbone ORB-SLAM2 system for real-time stereo shape recovery. Results: Experiments on in-vivo data sets show that BDIS-SLAM runs at over 30 Hz speed on modern single-core CPU in typical endoscopy/colonoscopy scenarios. BDIS-SLAM only consumes around an additional 12% time compared with the backbone ORB-SLAM2. Although our lightweight BDIS-SLAM simplifies the process by ignoring deformation and fusion procedures, it can provide a usable dense mapping for modern MIS on computationally constrained devices. Conclusion: The proposed BDIS-SLAM is a lightweight stereo dense SLAM system for MIS. It achieves 30 Hz on a modern single-core CPU in typical endoscopy/colonoscopy scenarios (image size around 640*480). BDIS-SLAM provides a low-cost solution for dense mapping in MIS and has the potential to be applied in surgical robots and AR systems., Comment: This paper has been accepted by International Journal of Computer Assisted Radiology and Surgery. Code is available at https://github.com/JingweiSong/BDIS-SLAM

Taking the nonextensive Tsallis and R\'enyi entropies into account, we explore thermodynamic properties and phase transitions of the Kerr-Newman black holes (KNBH) in the microcanonical and canonical ensembles. We also compare our results with those obtained by attributing the Bekenstein-Hawking entropy bound to the mentioned black holes. Our analysis indicates that, similarly to the standard Boltzmann picture, isolated KNBH in the microcanonical approach are stable against axisymmetric perturbations in both Tsallis and R\'enyi models. On the other hand, in considering the case when the black holes are enveloped by a bath of thermal radiation in the canonical treatment, the KNBH based on the Tsallis and R\'enyi entropies can be stable for some values of the entropy parameters, in contrast to the traditional Boltzmann framework. For the case of R\'enyi entropy, we find that a Hawking-Page transition and a first order small black hole/large black hole transition can occur in a similar fashion as in rotating black holes in an anti-de Sitter space. Finally, we employ the Ruppeneir geometrothermodynamic technique to provide a new perspective on studying the nature of interactions between black hole microstructures, revealing a non-trivial impact of nonextensive entropies., Comment: 15 pages, 17 figures

We present an open-source Python library for building and using datasets where inputs are clusters of textual data, and outputs are sequences of real values representing one or more time series signals. The news-signals library supports diverse data science and NLP problem settings related to the prediction of time series behaviour using textual data feeds. For example, in the news domain, inputs are document clusters corresponding to daily news articles about a particular entity, and targets are explicitly associated real-valued time series: the volume of news about a particular person or company, or the number of pageviews of specific Wikimedia pages. Despite many industry and research use cases for this class of problem settings, to the best of our knowledge, News Signals is the only open-source library designed specifically to facilitate data science and research settings with natural language inputs and time series targets. In addition to the core codebase for building and interacting with datasets, we also conduct a suite of experiments using several popular Machine Learning libraries, which are used to establish baselines for time series anomaly prediction using textual inputs., Comment: EMNLP NLP-OSS Workshop, December 2023

Low-precision fine-tuning of language models has gained prominence as a cost-effective and energy-efficient approach to deploying large-scale models in various applications. However, this approach is susceptible to the existence of outlier values in activation. The outlier values in the activation can negatively affect the performance of fine-tuning language models in the low-precision regime since they affect the scaling factor and thus make representing smaller values harder. This paper investigates techniques for mitigating outlier activation in low-precision integer fine-tuning of the language models. Our proposed novel approach enables us to represent the outlier activation values in 8-bit integers instead of floating-point (FP16) values. The benefit of using integers for outlier values is that it enables us to use operator tiling to avoid performing 16-bit integer matrix multiplication to address this problem effectively. We provide theoretical analysis and supporting experiments to demonstrate the effectiveness of our approach in improving the robustness and performance of low-precision fine-tuned language models.

Calibration of pixel values is a fundamental step for accurate measurements in astronomical imaging. In astronomical jargon this is known as estimating zero point magnitude. Here, we introduce a newly added script in GNU Astronomy Utilities (Gnuastro) version 0.20 for the zero point magnitude estimation, named: astscript-zeropoint. The script offers numerous features, such as the flexibility to use either image(s) or a catalog as the reference dataset. Additionally, steps are parallelized to enhance efficiency for big data. Thanks to Gnuastro's minimal dependencies, the script is both flexible and portable. The figures of this research note are reproducible with Maneage, on the Git commit c89275e., Comment: Supplementary data on Zenodo (https://doi.org/10.5281/zenodo.10256845), project source on Codeberg (https://codeberg.org/gnuastro/papers/src/branch/zeropoint) and archived on Software Heritage (swh:1:dir:8b2d1f63be96de3de03aa3e2bb68fa7fa52df56f; see text for usage)

We present an efficient parallel derandomization method for randomized algorithms that rely on concentrations such as the Chernoff bound. This settles a classic problem in parallel derandomization, which dates back to the 1980s. Consider the \textit{set balancing} problem where $m$ sets of size at most $s$ are given in a ground set of size $n$, and we should partition the ground set into two parts such that each set is split evenly up to a small additive (discrepancy) bound. A random partition achieves a discrepancy of $O(\sqrt{s \log m})$ in each set, by Chernoff bound. We give a deterministic parallel algorithm that matches this bound, using near-linear work and polylogarithmic depth. The previous results were weaker in discrepancy and/or work bounds: Motwani, Naor, and Naor [FOCS'89] and Berger and Rompel [FOCS'89] achieve discrepancy $s^{\varepsilon} \cdot O(\sqrt{s \log m})$ with work $\tilde{O}(m+n+\sum_{i=1}^{m} |S_i|) \cdot m^{\Theta(1/\varepsilon)}$ and polylogarithmic depth; the discrepancy was optimized to $O(\sqrt{s \log m})$ in later work, e.g. by Harris [Algorithmica'19], but the work bound remained high at $\tilde{O}(m^4n^3)$. Ghaffari, Grunau, and Rozhon [FOCS'23] achieve discrepancy $s/poly(\log(nm)) + O(\sqrt{s \log m})$ with near-linear work and polylogarithmic-depth. Notice that this discrepancy is barely sublinear with respect to the trivial bound of $s$. Our method relies on a novel bootstrapping idea that uses crude partitioning algorithms as a subroutine. In particular, we solve the problem recursively, by using the crude partition in each iteration to split the variables into many smaller parts, and then we find a constraint for the variables in each part such that we reduce the overall number of variables in the problem. The scheme relies on an interesting application of the multiplicative weights update method to control the variance losses in each iteration.

We present a novel technique for work-efficient parallel derandomization, for algorithms that rely on the concentration of measure bounds such as Chernoff, Hoeffding, and Bernstein inequalities. Our method increases the algorithm's computational work and depth by only polylogarithmic factors. Before our work, the only known method to obtain parallel derandomization with such strong concentrations was by the results of [Motwani, Naor, and Naor FOCS'89; Berger and Rompel FOCS'89], which perform a binary search in a $k$-wise independent space for $k=poly(\log n)$. However, that method blows up the computational work by a high $poly(n)$ factor and does not yield work-efficient parallel algorithms. Their method was an extension of the approach of [Luby FOCS'88], which gave a work-efficient derandomization but was limited to algorithms analyzed with only pairwise independence. Pushing the method from pairwise to the higher $k$-wise analysis resulted in the $poly(n)$ factor computational work blow-up. Our work can be viewed as an alternative extension from the pairwise case, which yields the desired strong concentrations while retaining work efficiency up to logarithmic factors. Our approach works by casting the problem of determining the random variables as an iterative process with $poly(\log n)$ iterations, where different iterations have independent randomness. This is done so that for the desired concentrations, we need only pairwise independence inside each iteration. In particular, we model each binary random variable as a result of a gradual random walk, and our method shows that the desired Chernoff-like concentrations about the endpoints of these walks can be boiled down to some pairwise analysis on the steps of these random walks in each iteration (while having independence across iterations).

This paper reports on developing a real-time invariant proprioceptive robot state estimation framework called DRIFT. A didactic introduction to invariant Kalman filtering is provided to make this cutting-edge symmetry-preserving approach accessible to a broader range of robotics applications. Furthermore, this work dives into the development of a proprioceptive state estimation framework for dead reckoning that only consumes data from an onboard inertial measurement unit and kinematics of the robot, with two optional modules, a contact estimator and a gyro filter for low-cost robots, enabling a significant capability on a variety of robotics platforms to track the robot's state over long trajectories in the absence of perceptual data. Extensive real-world experiments using a legged robot, an indoor wheeled robot, a field robot, and a full-size vehicle, as well as simulation results with a marine robot, are provided to understand the limits of DRIFT.