Search

Fault-tolerance is critically important in highly-distributed modern cloud applications. Solutions such as Temporal, Azure Durable Functions, and Beldi hide fault-tolerance complexity from developers by persisting execution state and resuming seamlessly from persisted state after failure. This pattern, often called durable execution, usually forces frequent and synchronous persistence and results in hefty latency overheads. In this paper, we propose distributed speculative execution (DSE), a technique for implementing the durable execution abstraction without incurring this penalty. With DSE, developers write code assuming synchronous persistence, and a DSE runtime is responsible for transparently bypassing persistence and reactively repairing application state on failure. We present libDSE, the first DSE application framework that achieves this vision. The key tension in designing libDSE is between imposing restrictions on user programs so the framework can safely and transparently change execution behavior, and avoiding assumptions so libDSE can support more use cases. We address this with a novel programming model centered around message-passing, atomic code blocks, and lightweight threads, and show that it allows developers to build a variety of speculative services, including write-ahead logs, key-value stores, event brokers, and fault-tolerant workflows. Our evaluation shows that libDSE reduces end-to-end latency by up to an order of magnitude compared to current generations of durable execution systems with minimal run-time overhead and manageable complexity.

This paper demonstrates numerically and experimentally that it is possible to tailor flexural band gaps in the low-frequency regime by appropriate choice of cutout characteristics. The finite element method is used to obtain the numerical dispersion relation and band gaps. The influence of the cutout's shape, size, and location on the band gap is systematically studied. The study demonstrates that the cutout should pass through the center of the unit cell, and a large aspect ratio is required to introduce flexural band gaps in the low-frequency regime. This is validated by experiments on a finite plate with 3 $\times$ 3 unit cells.

Mobile communications have been undergoing a generational change every ten years. While 5G network deployments are maturing, significant efforts are being made to standardize 6G, which is expected to be commercially introduced by 2030. This paper provides unique perspectives on the 6G network (radio and core) architecture(s) from the anticipated 6G use cases to meet the necessary performance requirements. To cater for the key 6G use cases, the 6G architecture must integrate different network-level functions in a multiplicity of virtual cloud environments, leveraging the advancements of distributed processing, artificial intelligence, and securely integrating different sub-networks e.g., terrestrial, and non-terrestrial networks into the overall 6G network. This paper characterizes the impact of 6G architectures from a deployment perspective with backwards compatibility in mind., Comment: 7 pages, 4 figures, accepted for publication in the IEEE Wireless Communications Magazine. arXiv admin note: text overlap with arXiv:2210.03286

Modern transformer-based encoder-decoder architectures struggle with reasoning tasks due to their inability to effectively extract relational information between input objects (data/tokens). Recent work introduced the Abstractor module, embedded between transformer layers, to address this gap. However, the Abstractor layer while excelling at capturing relational information (pure relational reasoning), faces challenges in tasks that require both object and relational-level reasoning (partial relational reasoning). To address this, we propose RESOLVE, a neuro-vector symbolic architecture that combines object-level features with relational representations in high-dimensional spaces, using fast and efficient operations such as bundling (summation) and binding (Hadamard product) allowing both object-level features and relational representations to coexist within the same structure without interfering with one another. RESOLVE is driven by a novel attention mechanism that operates in a bipolar high dimensional space, allowing fast attention score computation compared to the state-of-the-art. By leveraging this design, the model achieves both low compute latency and memory efficiency. RESOLVE also offers better generalizability while achieving higher accuracy in purely relational reasoning tasks such as sorting as well as partial relational reasoning tasks such as math problem-solving compared to state-of-the-art methods.

Self-dual Yang-Mills and Einstein gravity in Euclidean AdS$_4$ are useful toy models because they can be described by simple scalar Lagrangians exhibiting a new manifestation of the colour/kinematics duality, as recently shown by two of the authors. In this paper, we clarify how the self-dual sectors fit into the full theories. In particular, we explicitly construct the light-cone action for Yang-Mills theory and Einstein gravity in AdS$_4$ in terms of positive and negative helicity fields, where we are able to pinpoint the self-dual sector as expected. We then show that the boundary correlators of these theories take a remarkably simple form in terms of Feynman diagrams in half of flat space, acted on by certain differential operators. We also analyse their soft limits and show that they exhibit Weinberg-like soft factors, where the soft pole which appears in scattering amplitudes is replaced by a derivative with respect to the energy., Comment: 40 pages + 17 pages of appendix

Deep Learning methods are notorious for relying on extensive labeled datasets to train and assess their performance. This can cause difficulties in practical situations where models should be trained for new applications for which very little data is available. While few-shot learning algorithms can address the first problem, they still lack sufficient explanations for the results. This research presents a workflow that tackles both challenges by proposing an explainable few-shot learning workflow for detecting invasive and exotic tree species in the Atlantic Forest of Brazil using Unmanned Aerial Vehicle (UAV) images. By integrating a Siamese network with explainable AI (XAI), the workflow enables the classification of tree species with minimal labeled data while providing visual, case-based explanations for the predictions. Results demonstrate the effectiveness of the proposed workflow in identifying new tree species, even in data-scarce conditions. With a lightweight backbone, e.g., MobileNet, it achieves a F1-score of 0.86 in 3-shot learning, outperforming a shallow CNN. A set of explanation metrics, i.e., correctness, continuity, and contrastivity, accompanied by visual cases, provide further insights about the prediction results. This approach opens new avenues for using AI and UAVs in forest management and biodiversity conservation, particularly concerning rare or under-studied species.

Tests of general relativity (GR) can be systematically biased when our waveform models are inaccurate. We here study systematic biases in tests of general relativity induced by neglecting lensing effects for millilensed gravitational-wave signals, where the lens mass is typically in the $10^3M_\odot$--$10^5M_\odot$ range. In particular, we use a nested-sampling Bayesian parameter estimation and model selection analysis of a millilensed signal with an unlensed parameterized post-Einsteinian (ppE) recovery model. We find that the ppE model is significantly biased toward a detection of a deviation from general relativity at signal-to-noise ratios of 30 and higher, especially when the source is aligned with the lens mass (the lensing effect is pronounced) and when its total mass is low (the signal duration is long). We use a toy model and the linear signal and Laplace approximations to provide a semi-analytic explanation for the trends in the systematic errors found in the nested sampling analysis. Moreover, a Bayes factor analysis reveals that the (unlensed) ppE model is weakly favored over the (unlensed) GR model, and a fitting factor study shows there is a significant loss of signal-to-noise ratio when using the (unlensed) ppE model. This implies that although a parameter estimation study may incorrectly infer a deviation from general relativity, a residual signal-to-noise ratio test would reveal that the ppE model is not a good fit to the data. Thus, with current detectors, millilensing-induced systematic biases are unlikely to result in false positive detections of GR deviations., Comment: 25 pages, 14 figures

Our work aims to build a model that performs dual tasks of image captioning and image generation while being trained on only one task. The central idea is to train an invertible model that learns a one-to-one mapping between the image and text embeddings. Once the invertible model is efficiently trained on one task, the image captioning, the same model can generate new images for a given text through the inversion process, with no additional training. This paper proposes a simple invertible neural network architecture for this problem and presents our current findings., Comment: Accepted as Tiny Paper at ICVGIP 2024 conference

Despite the significant influx of prompt-tuning techniques for generative vision-language models (VLMs), it remains unclear how sensitive these models are to lexical and semantic alterations in prompts. In this paper, we evaluate the ability of generative VLMs to understand lexical and semantic changes in text using the SugarCrepe++ dataset. We analyze the sensitivity of VLMs to lexical alterations in prompts without corresponding semantic changes. Our findings demonstrate that generative VLMs are highly sensitive to such alterations. Additionally, we show that this vulnerability affects the performance of techniques aimed at achieving consistency in their outputs.

We study the robustness of parametrized tests of General Relativity (GR) with gravitational waves due to waveform inaccuracy. In particular, we determine the properties of the signal (signal-to-noise ratio (SNR) and source parameters) such that we are led to falsely identify a GR deviation due to neglecting spin precession or higher models in the recovery model. To characterize the statistical significance of the biases, we compute the Bayes factor between the non-GR and GR models, and the fitting factor of the non-GR model. For highly-precessing, edge-on signals, we find that mismodeling the signal leads to a significant systematic bias in the recovery of the non-GR parameter, even at an SNR of 30. However, these biased inferences are characterized by a significant loss of SNR and a weak preference for the non-GR model (over the GR model). At a higher SNR, the biased inferences display a strong preference for the non-GR model (over the GR model) and a significant loss of SNR. For edge-on signals containing asymmetric masses, at an SNR of 30, we find that excluding higher modes does not impact the ppE tests as much as excluding spin precession. Our analysis, therefore, identifies the spin-precessing and mass-asymmetric systems for which parametrized tests of GR are robust. With a toy model and using the linear signal approximation, we illustrate these regimes of bias and characterize them by obtaining bounds on the ratio of systematic to statistical error and the effective cycles incurred due to mismodeling. As a by-product of our analysis, we explicitly connect different measures and techniques commonly used to estimate systematic errors -- linear-signal approximation, Laplace approximation, fitting factor, effective cycles, and Bayes factor -- that are generally applicable to all studies of systematic uncertainties in gravitational wave parameter estimation., Comment: 51 pages, 22 figures, 5 tables

Time series anomaly detection (TSAD) is an evolving area of research motivated by its critical applications, such as detecting seismic activity, sensor failures in industrial plants, predicting crashes in the stock market, and so on. Across domains, anomalies occur significantly less frequently than normal data, making the F1-score the most commonly adopted metric for anomaly detection. However, in the case of time series, it is not straightforward to use standard F1-score because of the dissociation between `time points' and `time events'. To accommodate this, anomaly predictions are adjusted, called as point adjustment (PA), before the $F_1$-score evaluation. However, these adjustments are heuristics-based, and biased towards true positive detection, resulting in over-estimated detector performance. In this work, we propose an alternative adjustment protocol called ``Balanced point adjustment'' (BA). It addresses the limitations of existing point adjustment methods and provides guarantees of fairness backed by axiomatic definitions of TSAD evaluation., Comment: 5 pages, 6 figures

Automation transformed various aspects of our human civilization, revolutionizing industries and streamlining processes. In the domain of scientific inquiry, automated approaches emerged as powerful tools, holding promise for accelerating discovery, enhancing reproducibility, and overcoming the traditional impediments to scientific progress. This article evaluates the scope of automation within scientific practice and assesses recent approaches. Furthermore, it discusses different perspectives to the following questions: Where do the greatest opportunities lie for automation in scientific practice?; What are the current bottlenecks of automating scientific practice?; and What are significant ethical and practical consequences of automating scientific practice? By discussing the motivations behind automated science, analyzing the hurdles encountered, and examining its implications, this article invites researchers, policymakers, and stakeholders to navigate the rapidly evolving frontier of automated scientific practice.

Activation functions are non-linearities in neural networks that allow them to learn complex mapping between inputs and outputs. Typical choices for activation functions are ReLU, Tanh, Sigmoid etc., where the choice generally depends on the application domain. In this work, we propose a framework/strategy that unifies several works on activation functions and theoretically explains the performance benefits of these works. We also propose novel techniques that originate from the framework and allow us to obtain ``extensions'' (i.e. special generalizations of a given neural network) of neural networks through operations on activation functions. We theoretically and empirically show that ``extensions'' of neural networks have performance benefits compared to vanilla neural networks with insignificant space and time complexity costs on standard test functions. We also show the benefits of neural network ``extensions'' in the time-series domain on real-world datasets.

Study of correlation functions in AdS/CFT and in-in correlators in de Sitter space often requires the computation of Witten diagrams. Due to the complexity of evaluating radial integrals for these correlators, several indirect approaches have been developed to simplify computations. However, in momentum space, these methods have been limited to fields with integer spin. In this paper, we formulate tools for evaluating Witten diagrams with spin$-\frac12$ fields in momentum space and discuss where they differ from the corresponding integer-spin analysis. We formulate our tools explicitly for massless fermions and present how appropriate Weight shifting operators with respect to the external kinematics can be used to obtain the generalization to fermions with integer mass. We apply these tools to loop Witten diagrams and also discuss their use for evaluating in-in correlators in dS. In cases where we can evaluate the loop integrals, we find their transcendentality is lower than the corresponding scalar field results. Further, we classify the nature of IR divergences encountered for interacting massive scalars and fermions. We also prove a novel Weinberg-like soft theorem for gauge fields coupled to matter in AdS and show that the universal terms in the leading soft factor are sensitive to the spin of the matter field. These generalize the recently discovered soft theorems for pure Yang-Mills to Yang-Mills with matter., Comment: 50 pages + 15 pages of appendix; v2: typos fixed and references added

We derive formulae for the soft limit of tree-level gluon and graviton correlators in Anti-de Sitter space, which arise from Feynman diagrams encoding the Weinberg soft theorems in flat space. Other types of diagrams can also contribute to the soft limit at leading order in the soft momentum, but have a different pole structure. We derive these results at four points using explicit formulae recently obtained from the cosmological bootstrap and double copy, and extend them to any multiplicity using bootstrap techniques in Mellin-momentum space., Comment: v2. published in JHEP

This extended report presents DDS, a novel disaggregated storage architecture enabled by emerging networking hardware, namely DPUs (Data Processing Units). DPUs can optimize the latency and CPU consumption of disaggregated storage servers. However, utilizing DPUs for DBMSs requires careful design of the network and storage paths and the interface exposed to the DBMS. To fully benefit from DPUs, DDS heavily uses DMA, zero-copy, and userspace I/O to minimize overhead when improving throughput. It also introduces an offload engine that eliminates host CPUs by executing client requests directly on the DPU. Adopting DDS' API requires minimal DBMS modification. Our experimental study and production system integration show promising results -- DDS achieves higher disaggregated storage throughput with an order of magnitude lower latency, and saves up to tens of CPU cores per storage server., Comment: This paper is an extended report of DDS, which has been accepted by VLDB 2024

We explore the recently proposed explainable acoustic neural embedding~(XANE) system that models the background acoustics of a speech signal in a non-intrusive manner. The XANE embeddings are used to estimate specific parameters related to the background acoustic properties of the signal which allows the embeddings to be explainable in terms of those parameters. We perform ablation studies on the XANE system and show that estimating all acoustic parameters jointly has an overall positive effect. Furthermore, we illustrate the value of XANE embeddings by performing clustering experiments on unseen test data and show that the proposed embeddings achieve a mean F1 score of 92\% for three different tasks, outperforming significantly the WavLM based signal embeddings and are complimentary to speaker embeddings., Comment: arXiv admin note: substantial text overlap with arXiv:2406.05199

Despite their remarkable successes, state-of-the-art large language models (LLMs), including vision-and-language models (VLMs) and unimodal language models (ULMs), fail to understand precise semantics. For example, semantically equivalent sentences expressed using different lexical compositions elicit diverging representations. The degree of this divergence and its impact on encoded semantics is not very well understood. In this paper, we introduce the SUGARCREPE++ dataset to analyze the sensitivity of VLMs and ULMs to lexical and semantic alterations. Each sample in SUGARCREPE++ dataset consists of an image and a corresponding triplet of captions: a pair of semantically equivalent but lexically different positive captions and one hard negative caption. This poses a 3-way semantic (in)equivalence problem to the language models. We comprehensively evaluate VLMs and ULMs that differ in architecture, pre-training objectives and datasets to benchmark the performance of SUGARCREPE++ dataset. Experimental results highlight the difficulties of VLMs in distinguishing between lexical and semantic variations, particularly in object attributes and spatial relations. Although VLMs with larger pre-training datasets, model sizes, and multiple pre-training objectives achieve better performance on SUGARCREPE++, there is a significant opportunity for improvement. We show that all the models which achieve better performance on compositionality datasets need not perform equally well on SUGARCREPE++, signifying that compositionality alone may not be sufficient for understanding semantic and lexical alterations. Given the importance of the property that the SUGARCREPE++ dataset targets, it serves as a new challenge to the vision-and-language community., Comment: Added the dataset link to the abstract

We present a novel method for extracting neural embeddings that model the background acoustics of a speech signal. The extracted embeddings are used to estimate specific parameters related to the background acoustic properties of the signal in a non-intrusive manner, which allows the embeddings to be explainable in terms of those parameters. We illustrate the value of these embeddings by performing clustering experiments on unseen test data and show that the proposed embeddings achieve a mean F1 score of 95.2\% for three different tasks, outperforming significantly the WavLM based signal embeddings. We also show that the proposed method can explain the embeddings by estimating 14 acoustic parameters characterizing the background acoustics, including reverberation and noise levels, overlapped speech detection, CODEC type detection and noise type detection with high accuracy and a real-time factor 17 times lower than an external baseline method.

Human cognition excels at symbolic reasoning, deducing abstract rules from limited samples. This has been explained using symbolic and connectionist approaches, inspiring the development of a neuro-symbolic architecture that combines both paradigms. In parallel, recent studies have proposed the use of a "relational bottleneck" that separates object-level features from abstract rules, allowing learning from limited amounts of data . While powerful, it is vulnerable to the curse of compositionality meaning that object representations with similar features tend to interfere with each other. In this paper, we leverage hyperdimensional computing, which is inherently robust to such interference to build a compositional architecture. We adapt the "relational bottleneck" strategy to a high-dimensional space, incorporating explicit vector binding operations between symbols and relational representations. Additionally, we design a novel high-dimensional attention mechanism that leverages this relational representation. Our system benefits from the low overhead of operations in hyperdimensional space, making it significantly more efficient than the state of the art when evaluated on a variety of test datasets, while maintaining higher or equal accuracy.

The recent proliferation of personal wireless communication devices is driving the need for multi-band frequency selective components including multiplexers and dual-band filters. This paper presents a simple technique for transforming a single-band bandpass filter (BPF) into a dual-band BPF. A second order (two-pole) single-band bandpass filter was chosen for this research, giving rise to a fourth order (four-pole) dual-band bandpass filter after the proposed filter transformation. Both filters were then implemented using the compact U-shaped microstrip resonator for improved device miniaturization. The proposed work features a centre frequency of 1.4 GHz for the single-band bandpass filter, with a span of 3.4% fractional bandwidth. The dual-band bandpass filter operates at 1.35 and 1.45 GHz. The design implementation employs the commercially available Rogers RT/Duroid 6010LM substrate, having a dissipation factor of 0.0023, dielectric constant of 10.7, diel thickness (h) of 1.27 mm, and top/bottom cladding of 35 microns. The results reported for the theoretical and practical designs show good agreement and improved performance when compared to similar research works in literature. The practical responses of the prototype dual-band BPF indicate a good return loss of better than 18 dB across both bands, and an insertion loss of better than 0.1 dB., Comment: 4 pages, 6 figures, 1 table

In this work, we propose a heuristic based open source solver for finding global solution to constrained derivative-free optimization (DFO) problems. Our solver named Global optimization using Surrogates for Derivative-free Optimization (GSDO) relies on surrogate approximation to the original problem. In the proposed algorithm, an initial feasible point is first generated. This point is subsequently used to generate well spaced feasible points for formulating better radial basis function based surrogate approximations to original objective and constraint functions. Finally, these surrogates are used to solve the derivative-free global optimization problems. The proposed solver is capable of handling quantifiable and nonquantifiable as well as relaxable and unrelaxable constraints. We compared the performance of proposed solver with state of the art solvers like Nonlinear Optimization by Mesh Adaptive Direct Search (NOMAD), differential evolution (DE) and Simplicial Homology Global Optimization (SHGO) on standard test problems. The numerical results clearly demonstrate that the performance of our method is competitive with respect to other solvers., Comment: 17 pages, 5 figures, 8 tables

Despite their remarkable successes, state-of-the-art language models face challenges in grasping certain important semantic details. This paper introduces the VISLA (Variance and Invariance to Semantic and Lexical Alterations) benchmark, designed to evaluate the semantic and lexical understanding of language models. VISLA presents a 3-way semantic (in)equivalence task with a triplet of sentences associated with an image, to evaluate both vision-language models (VLMs) and unimodal language models (ULMs). An evaluation involving 34 VLMs and 20 ULMs reveals surprising difficulties in distinguishing between lexical and semantic variations. Spatial semantics encoded by language models also appear to be highly sensitive to lexical information. Notably, text encoders of VLMs demonstrate greater sensitivity to semantic and lexical variations than unimodal text encoders. Our contributions include the unification of image-to-text and text-to-text retrieval tasks, an off-the-shelf evaluation without fine-tuning, and assessing LMs' semantic (in)variance in the presence of lexical alterations. The results highlight strengths and weaknesses across diverse vision and unimodal language models, contributing to a deeper understanding of their capabilities. % VISLA enables a rigorous evaluation, shedding light on language models' capabilities in handling semantic and lexical nuances. Data and code will be made available at https://github.com/Sri-Harsha/visla_benchmark.

Previous works on depression detection use datasets collected in similar environments to train and test the models. In practice, however, the train and test distributions cannot be guaranteed to be identical. Distribution shifts can be introduced due to variations such as recording environment (e.g., background noise) and demographics (e.g., gender, age, etc). Such distributional shifts can surprisingly lead to severe performance degradation of the depression detection models. In this paper, we analyze the application of test-time training (TTT) to improve robustness of models trained for depression detection. When compared to regular testing of the models, we find TTT can significantly improve the robustness of the model under a variety of distributional shifts introduced due to: (a) background-noise, (b) gender-bias, and (c) data collection and curation procedure (i.e., train and test samples are from separate datasets).

We demonstrate the controllable generation of distinct types of dispersive shock-waves emerging in a quantum droplet bearing environment with the aid of step-like initial conditions. Dispersive regularization of the ensuing hydrodynamic singularities occurs due to the competition between meanfield repulsion and attractive quantum fluctuations. This interplay delineates the dominance of defocusing (hyperbolic) and focusing (elliptic) hydrodynamic phenomena respectively being designated by real and imaginary speed of sound. Specifically, the symmetries of the extended Gross-Pitaevskii model lead to a three-parameter family, encompassing two densities and a relative velocity, of the underlying Riemann problem utilized herein. Surprisingly, dispersive shock waves persist across the hyperbolic-to-elliptic threshold, while a plethora of additional wave patterns arise, such as rarefaction waves, traveling dispersive shock waves, (anti)kinks and droplet wavetrains. The classification and characterization of these features is achieved by deploying Whitham modulation theory. Our results pave the way for unveiling a multitude of unexplored coherently propagating waveforms in such attractively interacting mixtures and should be detectable by current experiments., Comment: 15 pages, 9 figures

In this paper, we introduce the problem of zero-shot text-guided exploration of the solutions to open-domain image super-resolution. Our goal is to allow users to explore diverse, semantically accurate reconstructions that preserve data consistency with the low-resolution inputs for different large downsampling factors without explicitly training for these specific degradations. We propose two approaches for zero-shot text-guided super-resolution - i) modifying the generative process of text-to-image \textit{T2I} diffusion models to promote consistency with low-resolution inputs, and ii) incorporating language guidance into zero-shot diffusion-based restoration methods. We show that the proposed approaches result in diverse solutions that match the semantic meaning provided by the text prompt while preserving data consistency with the degraded inputs. We evaluate the proposed baselines for the task of extreme super-resolution and demonstrate advantages in terms of restoration quality, diversity, and explorability of solutions., Comment: CVPR 2024

We study the problem of symbolic music generation (e.g., generating piano rolls), with a technical focus on non-differentiable rule guidance. Musical rules are often expressed in symbolic form on note characteristics, such as note density or chord progression, many of which are non-differentiable which pose a challenge when using them for guided diffusion. We propose Stochastic Control Guidance (SCG), a novel guidance method that only requires forward evaluation of rule functions that can work with pre-trained diffusion models in a plug-and-play way, thus achieving training-free guidance for non-differentiable rules for the first time. Additionally, we introduce a latent diffusion architecture for symbolic music generation with high time resolution, which can be composed with SCG in a plug-and-play fashion. Compared to standard strong baselines in symbolic music generation, this framework demonstrates marked advancements in music quality and rule-based controllability, outperforming current state-of-the-art generators in a variety of settings. For detailed demonstrations, code and model checkpoints, please visit our project website: https://scg-rule-guided-music.github.io/., Comment: ICML 2024 (Oral)

Low dose computed tomography (CT) acquisition using reduced radiation or sparse angle measurements is recommended to decrease the harmful effects of X-ray radiation. Recent works successfully apply deep networks to the problem of low dose CT recovery on bench-mark datasets. However, their robustness needs a thorough evaluation before use in clinical settings. In this work, we evaluate the robustness of different deep learning approaches and classical methods for CT recovery. We show that deep networks, including model-based networks encouraging data consistency, are more susceptible to untargeted attacks. Surprisingly, we observe that data consistency is not heavily affected even for these poor quality reconstructions, motivating the need for better regularization for the networks. We demonstrate the feasibility of universal attacks and study attack transferability across different methods. We analyze robustness to attacks causing localized changes in clinically relevant regions. Both classical approaches and deep networks are affected by such attacks leading to changes in the visual appearance of localized lesions, for extremely small perturbations. As the resulting reconstructions have high data consistency with the original measurements, these localized attacks can be used to explore the solution space of the CT recovery problem., Comment: MIDL 2023

In recent years, both online and offline deep learning models have been developed for time series forecasting. However, offline deep forecasting models fail to adapt effectively to changes in time-series data, while online deep forecasting models are often expensive and have complex training procedures. In this paper, we reframe the online nonlinear time-series forecasting problem as one of linear hyperdimensional time-series forecasting. Nonlinear low-dimensional time-series data is mapped to high-dimensional (hyperdimensional) spaces for linear hyperdimensional prediction, allowing fast, efficient and lightweight online time-series forecasting. Our framework, TSF-HD, adapts to time-series distribution shifts using a novel co-training framework for its hyperdimensional mapping and its linear hyperdimensional predictor. TSF-HD is shown to outperform the state of the art, while having reduced inference latency, for both short-term and long-term time series forecasting. Our code is publicly available at http://github.com/tsfhd2024/tsf-hd.git

Supersymmetric elementary string states in the compactified heterotic string theory are described by small black holes that have zero area event horizon. In this paper we compute the supersymmetric index of such elementary string states using gravitational path integral. The dominant contribution to the path integral comes from an Euclidean rotating black hole solution of the supergravity theory with a finite area event horizon, but the logarithm of the index, computed from the saddle point, vanishes. Nevertheless we show that the solution is singular on certain subspaces of the horizon where higher derivative corrections can be important, and once the higher derivative corrections are taken into account the solution could yield a finite result for the logarithm of the index whose form agrees with the microscopic results up to an overall numerical constant. While the numerical constant is not determined in our analysis, we show that it is independent of the details of the compactification and even the number of non-compact dimensions, in agreement with the microscopic results., Comment: 26 pages, v2: small changes, version to appear in JHEP

We investigate cosmological correlators for conformally coupled $\phi^4$ theory in four-dimensional de Sitter space. These \textit{in-in} correlators differ from scattering amplitudes for massless particles in flat space due to the spacelike structure of future infinity in de Sitter. They also require a regularization which preserves de Sitter-invariance, which makes the flat space limit subtle to define at loop-level. Nevertheless we find that up to two loops, the \textit{in-in} correlators are structurally simpler than the wave function and have the same transcendentality as flat space amplitudes. Moreover, we show that their loop integrands can be recast in terms of flat space integrands and can be derived from a novel recursion relation., Comment: 58 pages. Ancillary file recursions.nb implementing the recursions for the integrands of section 3 of the paper. v2: updated version with the evaluation of the correlators of phi_-. Various clarifications and References updated

The gravitational waves (GWs) emitted by neutron star binaries probe the physics of matter at supra nuclear densities. During the late inspiral, tidal deformations raised on each star by the gravitational field of its companion depend crucially on the star's internal properties. The misalignment of a star's tidal bulge with its companion's gravitational field encodes the strength of internal dissipative processes, which imprint onto the phase of the gravitational waves emitted. We here analyze GW data from the GW170817 (binary neutron star) event detected by LIGO and Virgo and find the first constraint on the dissipative tidal deformability of a neutron star. From this constraint, \emph{assuming} a temperature profile for each star in the binary, we obtain an order of magnitude bound on the averaged bulk ($\zeta$) and shear ($\eta$) viscosity of each star during the inspiral.: $\zeta \lesssim 10^{31} \mathrm{g}\;\mathrm{cm}^{-1}\mathrm{s}^{-1}$ and $\eta \lesssim 10^{28} \mathrm{g}\;\mathrm{cm}^{-1}\mathrm{s}^{-1} $. We forecast that these bounds could be improved by two orders of magnitude with third-generation detectors, like Cosmic Explorer, using inspiral data. These constraints already inform nuclear physics models and motivate further theoretical work to better understand the interplay between viscosity and temperature in the late inspiral of neutron stars., Comment: 7 pages, 4 figures; edits in response to referee comments; minor typos corrected

In this work, we propose novel designs of cellular structures exhibiting unconventional in-plane actuation responses to external stimuli. We strategically introduce stimuli-responsive bilayer bridge films within conventional honeycombs to achieve the desired actuation. The films are incorporated such that, in response to an external field (thermal, electric, chemical, etc.), the bridge film bends out-of-plane, activating the honeycomb in the plane. The conventional out-of-plane deformation of the bridge film can lead to interesting and unconventional actuation in the plane. An analytical model of this coupled unit cell behaviour is developed using curved beam theory, and the model is validated against finite element simulations. Several applications of such designs are presented. Unit cell architectures exhibiting both positive and negative macroscopic actuation are proposed, and the criterion for achieving such actuation is derived analytically. Furthermore, we demonstrate that by altering the topology, unidirectional and bidirectional negative actuation can be achieved. We also propose designs that result in the negative actuation of the structure with both monotonically increasing and monotonically decreasing stimuli. Finally, by combining two macroscopic structures with positive and negative actuation, we design efficient actuators/sensors that bend in the plane in response to a stimulus., Comment: 13 pages, 8 figures

Reinforcement Learning (RL) is notoriously data-inefficient, which makes training on a real robot difficult. While model-based RL algorithms (world models) improve data-efficiency to some extent, they still require hours or days of interaction to learn skills. Recently, offline RL has been proposed as a framework for training RL policies on pre-existing datasets without any online interaction. However, constraining an algorithm to a fixed dataset induces a state-action distribution shift between training and inference, and limits its applicability to new tasks. In this work, we seek to get the best of both worlds: we consider the problem of pretraining a world model with offline data collected on a real robot, and then finetuning the model on online data collected by planning with the learned model. To mitigate extrapolation errors during online interaction, we propose to regularize the planner at test-time by balancing estimated returns and (epistemic) model uncertainty. We evaluate our method on a variety of visuo-motor control tasks in simulation and on a real robot, and find that our method enables few-shot finetuning to seen and unseen tasks even when offline data is limited. Videos, code, and data are available at https://yunhaifeng.com/FOWM ., Comment: CoRL 2023 Oral; Project website: https://yunhaifeng.com/FOWM