Search

Here, we report on a new class active particles capable of dynamically programmable motion powered by electricity. We have implemented physical principles that separate the propulsion and steering mechanisms of active motion using optically activated, patterned, photoresponsive semiconductor coatings on intricate microstructures. Our engineered microswimmer robots employ an induced-charge electro-phoresis (ICEP) mechanism to achieve linear motion and optically modulated electrokinetic propulsion (OMEP) for steering. Optical modulation is achieved by manipulating the polarizability of patterned ZnO semiconductor coating through exposure to light with wavelengths above its bandgap, exploiting the semiconductor's photoconductive properties. Unlike previous methods that rely on changing the direction of optical illumination or spatially controlling narrow optical beams, our approach achieves optical steering under uniform ambient illumination conditions, thereby greatly reducing the complexity of the optical system. The decoupling of propulsion and steering allows for the programming of micromotor trajectories in both open and closed-loop control modes. We anticipate that our findings will pave the way for efficient optically gated control of the trajectory of photoresponsive active particles. Furthermore, they will enable the selective manipulation of specific subgroups of engineered active microparticles with various semiconducting coatings having different band gaps.

The block tensor of trifocal tensors provides crucial geometric information on the three-view geometry of a scene. The underlying synchronization problem seeks to recover camera poses (locations and orientations up to a global transformation) from the block trifocal tensor. We establish an explicit Tucker factorization of this tensor, revealing a low multilinear rank of $(6,4,4)$ independent of the number of cameras under appropriate scaling conditions. We prove that this rank constraint provides sufficient information for camera recovery in the noiseless case. The constraint motivates a synchronization algorithm based on the higher-order singular value decomposition of the block trifocal tensor. Experimental comparisons with state-of-the-art global synchronization methods on real datasets demonstrate the potential of this algorithm for significantly improving location estimation accuracy. Overall this work suggests that higher-order interactions in synchronization problems can be exploited to improve performance, beyond the usual pairwise-based approaches., Comment: 31 pages, 4 figures

We propose an efficient protocol for preparing low energy states of arbitrary fermionic Hamiltonians on a noisy bosonic quantum simulator. This procedure involves performing adiabatic cooling by coupling the target system with a simulated bath. The bath is periodically monitored in order to extract entropy from the system. By fermionizing the simulated target system and the bath together, we allow individual fermionic excitations of the system to coherently hop to the bath sites. In this way, we achieve a cooling rate linearly proportional to the density of these excitations, despite the fact that they are non-local in terms of the bosonic degrees of freedom of the hardware. In particular, we show that certain topological phases, such as the chiral (non-Abelian) phase of the Kitaev honeycomb model can be prepared efficiently using our protocol. We find that our protocol performs favorably in the presence of noise, making it suitable for execution on near-term quantum devices., Comment: 13 pages, 8 figures

Recent studies investigated the challenge of assessing the strength of a given claim extracted from a dataset, particularly the claim's potential of being misleading and cherry-picked. We focus on claims that compare answers to an aggregate query posed on a view that selects tuples. The strength of a claim amounts to the question of how likely it is that the view is carefully chosen to support the claim, whereas less careful choices would lead to contradictory claims. We embark on the study of the reverse task that offers a complementary angle in the critical assessment of data-based claims: given a claim, find useful supporting views. The goal of this task is twofold. On the one hand, we aim to assist users in finding significant evidence of phenomena of interest. On the other hand, we wish to provide them with machinery to criticize or counter given claims by extracting evidence of opposing statements. To be effective, the supporting sub-population should be significant and defined by a ``natural'' view. We discuss several measures of naturalness and propose ways of extracting the best views under each measure (and combinations thereof). The main challenge is the computational cost, as na\"ive search is infeasible. We devise anytime algorithms that deploy two main steps: (1) a preliminary construction of a ranked list of attribute combinations that are assessed using fast-to-compute features, and (2) an efficient search for the actual views based on each attribute combination. We present a thorough experimental study that shows the effectiveness of our algorithms in terms of quality and execution cost. We also present a user study to assess the usefulness of the naturalness measures.

We present Jamba-1.5, new instruction-tuned large language models based on our Jamba architecture. Jamba is a hybrid Transformer-Mamba mixture of experts architecture, providing high throughput and low memory usage across context lengths, while retaining the same or better quality as Transformer models. We release two model sizes: Jamba-1.5-Large, with 94B active parameters, and Jamba-1.5-Mini, with 12B active parameters. Both models are fine-tuned for a variety of conversational and instruction-following capabilties, and have an effective context length of 256K tokens, the largest amongst open-weight models. To support cost-effective inference, we introduce ExpertsInt8, a novel quantization technique that allows fitting Jamba-1.5-Large on a machine with 8 80GB GPUs when processing 256K-token contexts without loss of quality. When evaluated on a battery of academic and chatbot benchmarks, Jamba-1.5 models achieve excellent results while providing high throughput and outperforming other open-weight models on long-context benchmarks. The model weights for both sizes are publicly available under the Jamba Open Model License and we release ExpertsInt8 as open source., Comment: Webpage: https://www.ai21.com/jamba

A sensitive optical diffractometry method is developed and utilized for advanced tomography of laser-induced air plasma formations. Using transverse diffractometry and Supergaussian plasma distribution modelling we extract the main parameters of the plasma being the plasma density, width and shape with 20 micrometer spatial resolution throughout the plasma formation. The experimentally recorded diffraction patterns fitted by the Supergaussian plasma model are found to capture unprecedentedly delicate traits in the evolution of the plasma from its effective birth and on. Key features in the spatial evolution of the plasma such as the 'escape position', the 'turning point' and the refocusing dynamics of the beam are identified and explored in details. Our work provides experimental and theoretical access into the highly nonlinear dynamics of laser-induced air plasma., Comment: Main text - 10 pages, Supplementary Information - 6 pages

Diffusion models have opened the path to a wide range of text-based image editing frameworks. However, these typically build on the multi-step nature of the diffusion backwards process, and adapting them to distilled, fast-sampling methods has proven surprisingly challenging. Here, we focus on a popular line of text-based editing frameworks - the ``edit-friendly'' DDPM-noise inversion approach. We analyze its application to fast sampling methods and categorize its failures into two classes: the appearance of visual artifacts, and insufficient editing strength. We trace the artifacts to mismatched noise statistics between inverted noises and the expected noise schedule, and suggest a shifted noise schedule which corrects for this offset. To increase editing strength, we propose a pseudo-guidance approach that efficiently increases the magnitude of edits without introducing new artifacts. All in all, our method enables text-based image editing with as few as three diffusion steps, while providing novel insights into the mechanisms behind popular text-based editing approaches., Comment: Project page: https://turboedit-paper.github.io/

The recent demonstration of laser excitation of the $\approx 8$ eV isomeric state of thorium-229 is a significant step towards a nuclear clock. The low excitation energy likely results from a cancellation between the contributions of the electromagnetic and strong forces. Physics beyond the Standard Model could disrupt this cancellation, highlighting nuclear clocks' sensitivity to new physics. Accurate predictions of the different contributions to nuclear transition energies, and therefore of the quantitative sensitivity of a nuclear clock, are challenging. We improve upon previous sensitivity estimates and assess a ''nightmare scenario'', where all binding energy differences are small and the new physics sensitivity is poor. A classical geometric model of thorium-229 suggests that fine-tuning is needed for such a scenario. We also propose a $d$-wave halo model, inspired by effective field theory. We show that it reproduces observations and suggests the ''nightmare scenario'' is unlikely. We find that the nuclear clock's sensitivity to variations in the effective fine structure constant is enhanced by a factor of order $10^4$. We finally propose auxiliary nuclear measurements to reduce uncertainties and further test the validity of the halo model., Comment: 12 pages, 4 figures

Attention-based mechanisms are widely used in machine learning, most prominently in transformers. However, hyperparameters such as the rank of the attention matrices and the number of heads are scaled nearly the same way in all realizations of this architecture, without theoretical justification. In this work we show that there are dramatic trade-offs between the rank and number of heads of the attention mechanism. Specifically, we present a simple and natural target function that can be represented using a single full-rank attention head for any context length, but that cannot be approximated by low-rank attention unless the number of heads is exponential in the embedding dimension, even for short context lengths. Moreover, we prove that, for short context lengths, adding depth allows the target to be approximated by low-rank attention. For long contexts, we conjecture that full-rank attention is necessary. Finally, we present experiments with off-the-shelf transformers that validate our theoretical findings.

The recent laser excitation of the low-lying Th-229 isomer transition is starting a revolution in ultralight dark matter searches. The enhanced sensitivity of this transition to the large class of dark matter models dominantly coupling to quarks and gluons will ultimately allow us to probe coupling strengths eight orders of magnitude smaller than the current bounds from optical atomic clocks, which are mainly sensitive to dark matter couplings to electrons and photons. We argue that, with increasing precision, observations of the Th-229 excitation spectrum will soon give world-leading constraints. Using data from the pioneering laser excitation of Th-229 by Tiedau et al. [Phys. Rev. Lett. 132, 182501 (2024)], we present a first dark matter search in the excitation spectrum. While the exclusion limits of our detailed study of the lineshape are still below the sensitivity of currently operating clock experiments, we project the measurement of Zhang et al. [arXiv:2406.18719 [physics.atom-ph]] to surpass it., Comment: 14 pages, 6 figures

Current methods for reconstructing training data from trained classifiers are restricted to very small models, limited training set sizes, and low-resolution images. Such restrictions hinder their applicability to real-world scenarios. In this paper, we present a novel approach enabling data reconstruction in realistic settings for models trained on high-resolution images. Our method adapts the reconstruction scheme of arXiv:2206.07758 to real-world scenarios -- specifically, targeting models trained via transfer learning over image embeddings of large pre-trained models like DINO-ViT and CLIP. Our work employs data reconstruction in the embedding space rather than in the image space, showcasing its applicability beyond visual data. Moreover, we introduce a novel clustering-based method to identify good reconstructions from thousands of candidates. This significantly improves on previous works that relied on knowledge of the training set to identify good reconstructed images. Our findings shed light on a potential privacy risk for data leakage from models trained using transfer learning.

Large language models based on the transformer architectures can solve highly complex tasks. But are there simple tasks that such models cannot solve? Here we focus on very simple counting tasks, that involve counting how many times a token in the vocabulary have appeared in a string. We show that if the dimension of the transformer state is linear in the context length, this task can be solved. However, the solution we propose does not scale beyond this limit, and we provide theoretical arguments for why it is likely impossible for a size limited transformer to implement this task. Our empirical results demonstrate the same phase-transition in performance, as anticipated by the theoretical argument. Our results demonstrate the importance of understanding how transformers can solve simple tasks.

Sustaining long-term user engagement with mobile health (mHealth) interventions while preserving their high efficacy remains an ongoing challenge in real-world well-being applications. To address this issue, we introduce a new algorithm, the Personalized, Context-Aware Recommender (PCAR), for intervention selection and evaluate its performance in a field experiment. In a four-week, in-the-wild experiment involving 29 parents of young children, we delivered personalized stress-reducing micro-interventions through a mobile chatbot. We assessed their impact on stress reduction using momentary stress level ecological momentary assessments (EMAs) before and after each intervention. Our findings demonstrate the superiority of PCAR intervention selection in enhancing the engagement and efficacy of mHealth micro-interventions to stress coping compared to random intervention selection and a control group that did not receive any intervention. Furthermore, we show that even brief, one-minute interventions can significantly reduce perceived stress levels (p=0.001). We observe that individuals are most receptive to one-minute interventions during transitional periods between activities, such as transitioning from afternoon activities to bedtime routines. Our study contributes to the literature by introducing a personalized context-aware intervention selection algorithm that improves engagement and efficacy of mHealth interventions, identifying key timing for stress interventions, and offering insights into mechanisms to improve stress coping.

Hyperbolic representations have shown remarkable efficacy in modeling inherent hierarchies and complexities within data structures. Hyperbolic neural networks have been commonly applied for learning such representations from data, but they often fall short in preserving the geometric structures of the original feature spaces. In response to this challenge, our work applies the Gromov-Wasserstein (GW) distance as a novel regularization mechanism within hyperbolic neural networks. The GW distance quantifies how well the original data structure is maintained after embedding the data in a hyperbolic space. Specifically, we explicitly treat the layers of the hyperbolic neural networks as a transport map and calculate the GW distance accordingly. We validate that the GW distance computed based on a training set well approximates the GW distance of the underlying data distribution. Our approach demonstrates consistent enhancements over current state-of-the-art methods across various tasks, including few-shot image classification, as well as semi-supervised graph link prediction and node classification., Comment: Accepted for ECCV 2024

Group synchronization plays a crucial role in global pipelines for Structure from Motion (SfM). Its formulation is nonconvex and it is faced with highly corrupted measurements. Cycle consistency has been effective in addressing these challenges. However, computationally efficient solutions are needed for cycles longer than three, especially in practical scenarios where 3-cycles are unavailable. To overcome this computational bottleneck, we propose an algorithm for group synchronization that leverages information from cycles of lengths ranging from three to six with a time complexity of order $O(n^3)$ (or $O(n^{2.373})$ when using a faster matrix multiplication algorithm). We establish non-trivial theory for this and related methods that achieves competitive sample complexity, assuming the uniform corruption model. To advocate the practical need for our method, we consider distributed group synchronization, which requires at least 4-cycles, and we illustrate state-of-the-art performance by our method in this context.

Algorithmic price collusion facilitated by artificial intelligence (AI) algorithms raises significant concerns. We examine how AI agents using Q-learning engage in tacit collusion in two-sided markets. Our experiments reveal that AI-driven platforms achieve higher collusion levels compared to Bertrand competition. Increased network externalities significantly enhance collusion, suggesting AI algorithms exploit them to maximize profits. Higher user heterogeneity or greater utility from outside options generally reduce collusion, while higher discount rates increase it. Tacit collusion remains feasible even at low discount rates. To mitigate collusive behavior and inform potential regulatory measures, we propose incorporating a penalty term in the Q-learning algorithm.

Current adversarial attacks for multi-class classifiers choose the target class for a given input naively, based on the classifier's confidence levels for various target classes. We present a novel adversarial targeting method, \textit{MALT - Mesoscopic Almost Linearity Targeting}, based on medium-scale almost linearity assumptions. Our attack wins over the current state of the art AutoAttack on the standard benchmark datasets CIFAR-100 and ImageNet and for a variety of robust models. In particular, our attack is \emph{five times faster} than AutoAttack, while successfully matching all of AutoAttack's successes and attacking additional samples that were previously out of reach. We then prove formally and demonstrate empirically that our targeting method, although inspired by linear predictors, also applies to standard non-linear models.

We are increasingly subjected to the power of AI authorities. As AI decisions become inescapable, entering domains such as healthcare, education, and law, we must confront a vital question: how can we ensure AI systems have the legitimacy necessary for effective governance? This essay argues that to secure AI legitimacy, we need methods that engage the public in designing and constraining AI systems, ensuring these technologies reflect the community's shared values. Constitutional AI, proposed by Anthropic, represents a step towards this goal, offering a model for democratic control of AI. However, while Constitutional AI's commitment to hardcoding explicit principles into AI models enhances transparency and accountability, it falls short in two crucial aspects: addressing the opacity of individual AI decisions and fostering genuine democratic legitimacy. To overcome these limitations, this essay proposes "Public Constitutional AI." This approach envisions a participatory process where diverse stakeholders, including ordinary citizens, deliberate on the principles guiding AI development. The resulting "AI Constitution" would carry the legitimacy of popular authorship, grounding AI governance in the public will. Furthermore, the essay proposes "AI Courts" to develop "AI case law," providing concrete examples for operationalizing constitutional principles in AI training. This evolving combination of constitutional principles and case law aims to make AI governance more responsive to public values. By grounding AI governance in deliberative democratic processes, Public Constitutional AI offers a path to imbue automated authorities with genuine democratic legitimacy, addressing the unique challenges posed by increasingly powerful AI systems while ensuring their alignment with the public interest.

Quantum channels represent a broad spectrum of operations crucial to quantum information theory, encompassing everything from the transmission of quantum information to the manipulation of various resources. In the domain of states, the concept of majorization serves as a fundamental tool for comparing the uncertainty inherent in both classical and quantum systems. This paper establishes a rigorous framework for assessing the uncertainty in both classical and quantum channels. By employing a specific class of superchannels, we introduce and elucidate three distinct approaches to channel majorization: constructive, axiomatic, and operational. Intriguingly, these methodologies converge to a consistent ordering. This convergence not only provides a robust basis for defining entropy functions for channels but also clarifies the interpretation of entropy in this broader context. Most notably, our findings reveal that any viable entropy function for quantum channels must assume negative values, thereby challenging traditional notions of entropy., Comment: 16 pages (main text) + 21 pages (appendix), 6 figures, comments are welcome

The past few years have produced a series of spectacular advances in the decoding of speech from brain activity. The engine of these advances has been the acquisition of labelled data, with increasingly large datasets acquired from single subjects. However, participants exhibit anatomical and other individual differences, and datasets use varied scanners and task designs. As a result, prior work has struggled to leverage data from multiple subjects, multiple datasets, multiple tasks, and unlabelled datasets. In turn, the field has not benefited from the rapidly growing number of open neural data repositories to exploit large-scale data and deep learning. To address this, we develop an initial set of neuroscience-inspired self-supervised objectives, together with a neural architecture, for representation learning from heterogeneous and unlabelled neural recordings. Experimental results show that representations learned with these objectives scale with data, generalise across subjects, datasets, and tasks, and are also learned faster than using only labelled data. In addition, we set new benchmarks for two foundational speech decoding tasks. Taken together, these methods now unlock the potential for training speech decoding models with orders of magnitude more existing data., Comment: 10 pages, 4 figures, under review

This paper proposes Bayesian Adaptive Trials (BAT) as both an efficient method to conduct trials and a unifying framework for evaluation social policy interventions, addressing limitations inherent in traditional methods such as Randomized Controlled Trials (RCT). Recognizing the crucial need for evidence-based approaches in public policy, the proposal aims to lower barriers to the adoption of evidence-based methods and align evaluation processes more closely with the dynamic nature of policy cycles. BATs, grounded in decision theory, offer a dynamic, ``learning as we go'' approach, enabling the integration of diverse information types and facilitating a continuous, iterative process of policy evaluation. BATs' adaptive nature is particularly advantageous in policy settings, allowing for more timely and context-sensitive decisions. Moreover, BATs' ability to value potential future information sources positions it as an optimal strategy for sequential data acquisition during policy implementation. While acknowledging the assumptions and models intrinsic to BATs, such as prior distributions and likelihood functions, the paper argues that these are advantageous for decision-makers in social policy, effectively merging the best features of various methodologies.

With the increasing prevalence of probabilistic programming languages, Hamiltonian Monte Carlo (HMC) has become the mainstay of applied Bayesian inference. However HMC still struggles to sample from densities with multiscale geometry: a large step size is needed to efficiently explore low curvature regions while a small step size is needed to accurately explore high curvature regions. We introduce the delayed rejection generalized HMC (DR-G-HMC) sampler that overcomes this challenge by employing dynamic step size selection, inspired by differential equation solvers. In a single sampling iteration, DR-G-HMC sequentially makes proposals with geometrically decreasing step sizes if necessary. This simulates Hamiltonian dynamics with increasing fidelity that, in high curvature regions, generates proposals with a higher chance of acceptance. DR-G-HMC also makes generalized HMC competitive by decreasing the number of rejections which otherwise cause inefficient backtracking and prevents directed movement. We present experiments to demonstrate that DR-G-HMC (1) correctly samples from multiscale densities, (2) makes generalized HMC methods competitive with the state of the art No-U-Turn sampler, and (3) is robust to tuning parameters., Comment: 9 pages, 5 figures

Mergers of compact objects (binary neutron stars, BNS, or neutron star-black hole, NSBH) with a substantial mass ratio ($q>1.5$) are expected to produce a mildly relativistic ejecta within $\sim20^\circ$ from the orbital plane. We present a semi-analytic approach to calculate the expected synchrotron emission observed from various viewing angles, along with the corresponding radio maps, that are produced by a collisionless shock driven by such ejecta into the interstellar medium. This method is supported by 2D numerical calculations of the full relativistic hydrodynamics with an agreement in the observed emission of $\sim30\%$. We consider a toroid ejecta with an opening angle of $15^\circ\leq\theta_ \text{open}\leq30^\circ$ and broken power-law mass distribution, $M(>\gamma\beta)\propto(\gamma\beta)^{-s}$ with $s=s_{\rm KN}$ at $\gamma\beta<\gamma_0\beta_0$ and $s=s_{\rm ft}$ at $\gamma\beta>\gamma_0\beta_0$ (where $\gamma$ is the Lorentz factor). While the peak flux is dimmer by a factor of $\sim$2-3, and the peak time remains roughly the same (within $20\%$), for various viewing angles compared to isotropic equivalent ejecta ($\theta_\text{open}=90^\circ$) considered in preceding papers, the radio maps are significantly different from the spherical case. The semi-analytic method can provide information on the ejecta geometry and viewing angle from future radio map observations and, consequently, constrain the ejection mechanism. Due to the larger ejected mass in NSBH merger, this late non-thermal signal can be observed to distances of $\lesssim 200$Mpc for typical parameter values., Comment: 10 pages, 8 figures, submitted to MNRAS

Graph Neural Networks (GNNs) have emerged as the predominant approach for learning over graph-structured data. However, most GNNs operate as black-box models and require post-hoc explanations, which may not suffice in high-stakes scenarios where transparency is crucial. In this paper, we present a GNN that is interpretable by design. Our model, Graph Neural Additive Network (GNAN), is a novel extension of the interpretable class of Generalized Additive Models, and can be visualized and fully understood by humans. GNAN is designed to be fully interpretable, allowing both global and local explanations at the feature and graph levels through direct visualization of the model. These visualizations describe the exact way the model uses the relationships between the target variable, the features, and the graph. We demonstrate the intelligibility of GNANs in a series of examples on different tasks and datasets. In addition, we show that the accuracy of GNAN is on par with black-box GNNs, making it suitable for critical applications where transparency is essential, alongside high accuracy.

Magic states are essential for achieving universal quantum computation. This study introduces a reversible framework for the manipulation of magic states in odd dimensions, delineating a necessary and sufficient condition for the exact transformations between magic states under maps that preserve the trace of states and positivity of discrete Wigner representation. Utilizing the stochastic formalism, we demonstrate that magic mana emerges as the unique measure for such reversible magic state transformations. We propose the concept of physical implementability for characterizing the hardness and cost of maintaining reversibility. Our findings show that, analogous to the entanglement theory, going beyond the positivity constraint enables an exact reversible theory of magic manipulation, thereby hinting at a potential incongruity between the reversibility of quantum resources and the fundamental principles of quantum mechanics. Physical implementability for reversible manipulation provides a new perspective for understanding and quantifying quantum resources, contributing to an operational framework for understanding the cost of reversible quantum resource manipulation., Comment: 14 pages including appendix

We show that, in the Nelson-Barr solution to the strong CP-problem, a naturally light scalar can arise. It gives rise to a completely new phenomenology beyond that of the celebrated QCD axion, if this field constitutes dark matter, as the CKM elements vary periodically in time. We also discuss how the model can be tested using quantum sensors, in particular using nuclear clocks, which leads to an interesting synergy between different frontiers of physics., Comment: 8 pages

Discounted algorithms often encounter evaluation errors due to their reliance on short-term estimations, which can impede their efficacy in addressing simple, short-term tasks and impose undesired temporal discounts (\(\gamma\)). Interestingly, these algorithms are often tested without applying a discount, a phenomenon we refer as the \textit{train-test bias}. In response to these challenges, we propose the Markov Flow Policy, which utilizes a non-negative neural network flow to enable comprehensive forward-view predictions. Through integration into the TD7 codebase and evaluation using the MuJoCo benchmark, we observe significant performance improvements, positioning MFP as a straightforward, practical, and easily implementable solution within the domain of average rewards algorithms., Comment: Paper has been not finished

We consider an effective theory for a shift-symmetric, quadratically-coupled, ultralight spin-0 field. The leading CP-conserving interactions with Standard Model fields in the effective theory arise at dimension 8. We discuss the renormalization group evolution and positivity bounds on these operators, as well as their possible UV origins. Assuming that the spin-0 field is associated with an ultralight dark matter candidate, we discuss the effects of the dimension-8 operators on experiments searching for the oscillation of fundamental constants and Lorentz violation. We find that the direct bounds on these two effects are of similar strength but rather weak, corresponding to a UV cutoff scale of keV order, as they are mediated by dimension-8 operators., Comment: 28 pages, 6 figures, 1 table; Edited acknowledgment; Edited for submission to JHEP

We present the subspace-constrained Tyler's estimator (STE) designed for recovering a low-dimensional subspace within a dataset that may be highly corrupted with outliers. STE is a fusion of the Tyler's M-estimator (TME) and a variant of the fast median subspace. Our theoretical analysis suggests that, under a common inlier-outlier model, STE can effectively recover the underlying subspace, even when it contains a smaller fraction of inliers relative to other methods in the field of robust subspace recovery. We apply STE in the context of Structure from Motion (SfM) in two ways: for robust estimation of the fundamental matrix and for the removal of outlying cameras, enhancing the robustness of the SfM pipeline. Numerical experiments confirm the state-of-the-art performance of our method in these applications. This research makes significant contributions to the field of robust subspace recovery, particularly in the context of computer vision and 3D reconstruction., Comment: 23 pages, accepted by CVPR 24

Effective surgical planning for breast cancer hinges on accurately predicting pathological complete response (pCR) to neoadjuvant chemotherapy (NAC). Diffusion-weighted MRI (DWI) and machine learning offer a non-invasive approach for early pCR assessment. However, most machine-learning models require manual tumor segmentation, a cumbersome and error-prone task. We propose a deep learning model employing "Size-Adaptive Lesion Weighting" for automatic DWI tumor segmentation to enhance pCR prediction accuracy. Despite histopathological changes during NAC complicating DWI image segmentation, our model demonstrates robust performance. Utilizing the BMMR2 challenge dataset, it matches human experts in pCR prediction pre-NAC with an area under the curve (AUC) of 0.76 vs. 0.796, and surpasses standard automated methods mid-NAC, with an AUC of 0.729 vs. 0.654 and 0.576. Our approach represents a significant advancement in automating breast cancer treatment planning, enabling more reliable pCR predictions without manual segmentation., Comment: Accepted for presentation at the IEEE International Symposium on Biomedical Imaging (ISBI)

We describe a system called Qr-Hint that, given a (correct) target query Q* and a (wrong) working query Q, both expressed in SQL, provides actionable hints for the user to fix the working query so that it becomes semantically equivalent to the target. It is particularly useful in an educational setting, where novices can receive help from Qr-Hint without requiring extensive personal tutoring. Since there are many different ways to write a correct query, we do not want to base our hints completely on how Q* is written; instead, starting with the user's own working query, Qr-Hint purposefully guides the user through a sequence of steps that provably lead to a correct query, which will be equivalent to Q* but may still "look" quite different from it. Ideally, we would like Qr-Hint's hints to lead to the "smallest" possible corrections to Q. However, optimality is not always achievable in this case due to some foundational hurdles such as the undecidability of SQL query equivalence and the complexity of logic minimization. Nonetheless, by carefully decomposing and formulating the problems and developing principled solutions, we are able to provide provably correct and locally optimal hints through Qr-Hint. We show the effectiveness of Qr-Hint through quality and performance experiments as well as a user study in an educational setting., Comment: SIGMOD 2024

The nature of dark matter (DM) and its interaction with the Standard Model (SM) is one of the biggest open questions in physics nowadays. The vast majority of theoretically-motivated Ultralight-DM (ULDM) models predict that ULDM couples dominantly to the SM strong/nuclear sector. This coupling leads to oscillations of nuclear parameters that are detectable by comparing clocks with different sensitivities to these nature's constants. Vibrational transitions of molecular clocks are more sensitive to a change in the nuclear parameters than the electronic transitions of atomic clocks. Here, we propose the iodine molecular ion, I$_2^+$, as a sensitive detector for such a class of ULDM models. The iodine's dense spectrum allows us to match its transition frequency to that of an optical atomic clock (Ca$^+$) and perform correlation spectroscopy between the two clock species. With this technique, we project a few-orders-of-magnitude improvement over the most sensitive clock comparisons performed to date. We also briefly consider the robustness of the corresponding "Earth-bound" under modifications of the $Z_N$-QCD axion model., Comment: 10 pages, 4 figures; v2 matches version published in PRD

This work analyzes the subspace-constrained Tyler's estimator (STE) designed for recovering a low-dimensional subspace within a dataset that may be highly corrupted with outliers. It assumes a weak inlier-outlier model and allows the fraction of inliers to be smaller than a fraction that leads to computational hardness of the robust subspace recovery problem. It shows that in this setting, if the initialization of STE, which is an iterative algorithm, satisfies a certain condition, then STE can effectively recover the underlying subspace. It further shows that under the generalized haystack model, STE initialized by the Tyler's M-estimator (TME), can recover the subspace when the fraction of iniliers is too small for TME to handle.

The rapid integration of Large Language Models (LLMs) across diverse sectors has marked a transformative era, showcasing remarkable capabilities in text generation and problem-solving tasks. However, this technological advancement is accompanied by significant risks and vulnerabilities. Despite ongoing security enhancements, attackers persistently exploit these weaknesses, casting doubts on the overall trustworthiness of LLMs. Compounding the issue, organisations are deploying LLM-integrated systems without understanding the severity of potential consequences. Existing studies by OWASP and MITRE offer a general overview of threats and vulnerabilities but lack a method for directly and succinctly analysing the risks for security practitioners, developers, and key decision-makers who are working with this novel technology. To address this gap, we propose a risk assessment process using tools like the OWASP risk rating methodology which is used for traditional systems. We conduct scenario analysis to identify potential threat agents and map the dependent system components against vulnerability factors. Through this analysis, we assess the likelihood of a cyberattack. Subsequently, we conduct a thorough impact analysis to derive a comprehensive threat matrix. We also map threats against three key stakeholder groups: developers engaged in model fine-tuning, application developers utilizing third-party APIs, and end users. The proposed threat matrix provides a holistic evaluation of LLM-related risks, enabling stakeholders to make informed decisions for effective mitigation strategies. Our outlined process serves as an actionable and comprehensive tool for security practitioners, offering insights for resource management and enhancing the overall system security., Comment: 10 pages, 1 figure, 3 tables

Meta-reinforcement learning (meta-RL) is a promising framework for tackling challenging domains requiring efficient exploration. Existing meta-RL algorithms are characterized by low sample efficiency, and mostly focus on low-dimensional task distributions. In parallel, model-based RL methods have been successful in solving partially observable MDPs, of which meta-RL is a special case. In this work, we leverage this success and propose a new model-based approach to meta-RL, based on elements from existing state-of-the-art model-based and meta-RL methods. We demonstrate the effectiveness of our approach on common meta-RL benchmark domains, attaining greater return with better sample efficiency (up to $15\times$) while requiring very little hyperparameter tuning. In addition, we validate our approach on a slate of more challenging, higher-dimensional domains, taking a step towards real-world generalizing agents.

Binary neutron star mergers are expected to produce fast dynamical ejecta, with mildly relativistic velocities extending to $\beta=v/c>0.6$. In a preceding paper, we derived an analytic description of the time-dependent radio to X-ray synchrotron flux produced by collisionless shocks driven by such fast ejecta into the interstellar medium, for spherical ejecta with broken power-law mass (or energy) distributions, $M(>\gamma\beta)\propto(\gamma\beta)^{-s}$ with $s=s_\text{KN}$ at $\gamma\beta<\gamma_0\beta_0$ and $s=s_\text{ft}$ at $\gamma\beta>\gamma_0\beta_0$ (where $\gamma$ is the Lorentz factor). Here, we extend our analysis and provide analytic expressions for the self-absorption frequency, the cooling frequency, and the observed angular size of the emitting region (which appears as a ring in the sky). For parameter values characteristic of merger calculation results -- a "shallow" mass distribution, $15$, "fast tail" mass distribution -- the analytic results reproduce well (to tens of percent accuracy) the results of detailed numeric calculations, a significant improvement over earlier order-of-magnitude estimates (based on extrapolations of results valid for $\gamma\beta\ll1$)., Comment: 8 pages, 7 figures, accepted to MNRAS

We investigate the training and generalization errors of overparameterized neural networks (NNs) with a wide class of leaky rectified linear unit (ReLU) functions. More specifically, we carefully upper bound both the convergence rate of the training error and the generalization error of such NNs and investigate the dependence of these bounds on the Leaky ReLU parameter, $\alpha$. We show that $\alpha =-1$, which corresponds to the absolute value activation function, is optimal for the training error bound. Furthermore, in special settings, it is also optimal for the generalization error bound. Numerical experiments empirically support the practical choices guided by the theory.

Deep Neural Nets (DNNs) learn latent representations induced by their downstream task, objective function, and other parameters. The quality of the learned representations impacts the DNN's generalization ability and the coherence of the emerging latent space. The Information Bottleneck (IB) provides a hypothetically optimal framework for data modeling, yet it is often intractable. Recent efforts combined DNNs with the IB by applying VAE-inspired variational methods to approximate bounds on mutual information, resulting in improved robustness to adversarial attacks. This work introduces a new and tighter variational bound for the IB, improving performance of previous IB-inspired DNNs. These advancements strengthen the case for the IB and its variational approximations as a data modeling framework, and provide a simple method to significantly enhance the adversarial robustness of classifier DNNs., Comment: 10 pages, 5 figures, code included in github repo

This book delves into the burgeoning field of quantum resource theories, a novel and vibrant area of research within quantum information science that seeks to unify diverse quantum phenomena under a single framework. By recognizing various attributes of physical systems as "resources," this approach offers a fresh perspective on quantum phenomena, transforming our understanding and application of concepts such as quantum entanglement, coherence, and more. With a focus on the pedagogical, the book aims to equip readers with the advanced mathematical tools and physical principles needed to navigate and contribute to this rapidly evolving field. It covers a wide range of topics, from the foundational aspects of quantum mechanics and quantum information to detailed explorations of specific resource theories, including entanglement, asymmetry, and thermodynamics. Through rigorous mathematical exposition and a unique axiomatic approach, the book provides deep insights into the operational and conceptual frameworks that underpin quantum resource theories, making it an invaluable resource for graduate students, early-career researchers, and anyone interested in the cutting-edge developments in quantum information science., Comment: 956 Pages (including appendices), Preliminary Version, Feedback and comments are most welcome especially typos, errors, and missing references

Set reconciliation, where two parties hold fixed-length bit strings and run a protocol to learn the strings they are missing from each other, is a fundamental task in many distributed systems. We present Rateless Invertible Bloom Lookup Tables (Rateless IBLT), the first set reconciliation protocol, to the best of our knowledge, that achieves low computation cost and near-optimal communication cost across a wide range of scenarios: set differences of one to millions, bit strings of a few bytes to megabytes, and workloads injected by potential adversaries. Rateless IBLT is based on a novel encoder that incrementally encodes the set difference into an infinite stream of coded symbols, resembling rateless error-correcting codes. We compare Rateless IBLT with state-of-the-art set reconciliation schemes and demonstrate significant improvements. Rateless IBLT achieves 3--4x lower communication cost than non-rateless schemes with similar computation cost, and 2--2000x lower computation cost than schemes with similar communication cost. We show the real-world benefits of Rateless IBLT by applying it to synchronize the state of the Ethereum blockchain, and demonstrate 5.6x lower end-to-end completion time and 4.4x lower communication cost compared to the system used in production., Comment: SIGCOMM 2024