Search

A broad range of technologies rely on remote inference, wherein data acquired is conveyed over a communication channel for inference in a remote server. Communication between the participating entities is often carried out over rate-limited channels, necessitating data compression for reducing latency. While deep learning facilitates joint design of the compression mapping along with encoding and inference rules, existing learned compression mechanisms are static, and struggle in adapting their resolution to changes in channel conditions and to dynamic links. To address this, we propose Adaptive Rate Task-Oriented Vector Quantization (ARTOVeQ), a learned compression mechanism that is tailored for remote inference over dynamic links. ARTOVeQ is based on designing nested codebooks along with a learning algorithm employing progressive learning. We show that ARTOVeQ extends to support low-latency inference that is gradually refined via successive refinement principles, and that it enables the simultaneous usage of multiple resolutions when conveying high-dimensional data. Numerical results demonstrate that the proposed scheme yields remote deep inference that operates with multiple rates, supports a broad range of bit budgets, and facilitates rapid inference that gradually improves with more bits exchanged, while approaching the performance of single-rate deep quantization methods., Comment: 13 pages, 12 figures

With today's multi-petawatt lasers, testing quantum electrodynamics (QED) in the strong field regime, where the electric field exceeds the Schwinger critical field in the rest frame of an electron, becomes within reach. Inverse Compton scattering of an intense laser pulse off a high-energy electron beam is the mainstream approach, resulting in the emission of high-energy photons that can decay into Breit-Wheeler electron-positron pairs. Here, we demonstrate experimentally that very high energy photons can be generated in a self-aligned single-laser Compton scattering setup, combining a laser-plasma accelerator and a plasma mirror. Reaching up to the GeV scale, photon emission via nonlinear Compton scattering exhibits a nonclassical scaling in the experiment that is consistent with electric fields reaching up to a fraction $\chi\simeq0.3$ of the Schwinger field in the electron rest frame. These foolproof collisions guaranteed by automatic laser-electron overlap provide a new approach for precise investigations of strong-field QED processes.

In 1972, Branko Gr\"unbaum conjectured that any arrangement of $n>2$ pairwise crossing pseudocircles in the plane can have at most $2n-2$ digons (regions enclosed by exactly two pseudoarcs), with the bound being tight. While this conjecture has been confirmed for cylindrical arrangements of pseudocircles and more recently for geometric circles, we extend these results to any simple arrangement of pairwise intersecting pseudocircles. Using techniques from the above-mentioned special cases, we provide a complete proof of Gr\"unbaum's conjecture that has stood open for over five decades.

We present a scalable, robust approach to creating quantum programs of arbitrary size and complexity. The approach is based on the true abstraction of the problem. The quantum program is expressed in terms of a high-level model together with constraints and objectives on the final program. Advanced synthesis algorithms transform the model into a low-level quantum program that meets the user's specification and is directed at a stipulated hardware. This separation of description from implementation is essential for scale. The technology adapts electronic design automation methods to quantum computing, finding feasible implementations in a virtually unlimited functional space. The results show clear superiority over the compilation and transpilation methods used today. We expect that this technological approach will take over and prevail as quantum software become more demanding, complex, and essential.

We present a novel method for 3D scene editing using diffusion models, designed to ensure view consistency and realism across perspectives. Our approach leverages attention features extracted from a single reference image to define the intended edits. These features are warped across multiple views by aligning them with scene geometry derived from Gaussian splatting depth estimates. Injecting these warped features into other viewpoints enables coherent propagation of edits, achieving high fidelity and spatial alignment in 3D space. Extensive evaluations demonstrate the effectiveness of our method in generating versatile edits of 3D scenes, significantly advancing the capabilities of scene manipulation compared to the existing methods. Project page: \url{https://attention-warp.github.io}

Generative AI is transforming enterprise application development by enabling machines to create content, code, and designs. These models, however, demand substantial computational power and data management. Cloud computing addresses these needs by offering infrastructure to train, deploy, and scale generative AI models. This review examines cloud services for generative AI, focusing on key providers like Amazon Web Services (AWS), Microsoft Azure, Google Cloud, IBM Cloud, Oracle Cloud, and Alibaba Cloud. It compares their strengths, weaknesses, and impact on enterprise growth. We explore the role of high-performance computing (HPC), serverless architectures, edge computing, and storage in supporting generative AI. We also highlight the significance of data management, networking, and AI-specific tools in building and deploying these models. Additionally, the review addresses security concerns, including data privacy, compliance, and AI model protection. It assesses the performance and cost efficiency of various cloud providers and presents case studies from healthcare, finance, and entertainment. We conclude by discussing challenges and future directions, such as technical hurdles, vendor lock-in, sustainability, and regulatory issues. Put together, this work can serve as a guide for practitioners and researchers looking to adopt cloud-based generative AI solutions, serving as a valuable guide to navigating the intricacies of this evolving field., Comment: 65 pages, 10 Figures,and supplementary methods detailing extended technical descriptions, service matrices, SWOT analyses, and detailed provider comparisons

We construct a model of the form $L[A,U]$ that exhibits the simplest structural behavior of $\sigma$-complete ultrafilters in a model of set theory with a single measurable cardinal $\kappa$ , yet satisfies $2^\kappa = \kappa^{++}$. This result establishes a limitation on the extent to which structural properties of ultrafilters can determine the cardinal arithmetic at large cardinals, and answers a question posed by Goldberg concerning the failure of the Continuum Hypothesis at a measurable cardinal in a model of the Ultrapower Axiom. The construction introduces several methods in extensions of embeddings theory and fine-structure-based forcing, designed to control the behavior of non-normal ultrafilters in generic extensions.

Nanowires (NWs) of topological materials are emerging as an exciting platform to probe and engineer new quantum phenomena that are hard to access in bulk phase. Their quasi-one-dimensional geometry and large surface-to-bulk ratio unlock new expressions of topology and highlight surface states. TaAs2, a compensated semimetal, is a topologically rich material harboring nodal-line, weak topological insulator (WTI), C2-protected topological crystalline insulator, and Zeeman field-induced Weyl semimetal phases. We report the synthesis of TaAs2 NWs in situ encapsulated in a dielectric SiO2 shell, which enabled us to probe rich magnetotransport phenomena, including metal-to-insulator transition and strong signatures of topologically non-trivial transport at remarkably high temperatures, direction-dependent giant positive and negative magnetoresistance, and a double pattern of Aharonov-Bohm oscillations, demonstrating coherent surface transport consistent with the two Dirac cones of a WTI surface. The coexistence and susceptibility of topological phases to external stimuli have potential applications in spintronics and nanoscale quantum technology., Comment: 23 PAGES, 04 FIGURES

Understanding the dynamics of an epidemic spread is crucial for effective control measures. During the COVID-19 pandemic, quarantines were implemented to minimize infections while mitigating social and economic impacts, raising the question of how to maximize quarantine efficiency. Previous research on periodic quarantines using the susceptible-infected-recovered (SIR) and similar models identified optimal duration for periodic quarantines. However, the question of the optimal initiation time for a single quarantine remains unanswered. Here, we use the SIR model in order to determine the optimal quarantine initiation time, by computing the optimal susceptible fraction at the onset of the quarantine, which minimizes the total outbreak size. Our analysis extends from a well-mixed scenario to strongly-heterogeneous social networks. We show that the optimal quarantine initiation time is closely related to the so-called "herd immunity" threshold, occurring at the onset of epidemic decline. Importantly, providing a methodology for identifying the optimal quarantine initiation time across different network structures, entails significant implications for epidemic control., Comment: 10 pages, 6 figures

We propose an extended demand response program, based on ancillary service for supplying flexible electricity demand. In our proposed scheme, we suggest a broader management model to control the scheduling and power consumption of Bitcoin mining machines. The main aspect that we focus on is suppressing the power ramping and related transient effects. We extend previous works on the subject, that study the impact of incorporating cryptocurrency mining machines into existing power grid, and explore the potential profit of exploiting this flexible load in the Israeli electricity market. We analyze a trend based on historical data, of increasing electricity prices and ramping costs due to the increasing penetration of renewable energy sources. We suggest an extension to the unit commitment problem from which we obtain the scheduling scheme of the Bitcoin mining machines. We use simulation and the real-world data acquired from the "Noga" grid operator to verify the proposed ancillary service and test its practical limits for reducing the ramping costs, under changing ratio of energy production from renewable sources. Out results suggests that the machine price and ratio of production from renewable sources plays a significant role in determining the profitability of the proposed demand-response program., Comment: 8 pages, 7 figures

We study finite-player dynamic stochastic games with heterogeneous interactions and non-Markovian linear-quadratic objective functionals. We derive the Nash equilibrium explicitly by converting the first-order conditions into a coupled system of stochastic Fredholm equations, which we solve in terms of operator resolvents. When the agents' interactions are modeled by a weighted graph, we formulate the corresponding non-Markovian continuum-agent game, where interactions are modeled by a graphon. We also derive the Nash equilibrium of the graphon game explicitly by first reducing the first-order conditions to an infinite-dimensional coupled system of stochastic Fredholm equations, then decoupling it using the spectral decomposition of the graphon operator, and finally solving it in terms of operator resolvents. Moreover, we show that the Nash equilibria of finite-player games on graphs converge to those of the graphon game as the number of agents increases. This holds both when a given graph sequence converges to the graphon in the cut norm and when the graph sequence is sampled from the graphon. We also bound the convergence rate, which depends on the cut norm in the former case and on the sampling method in the latter. Finally, we apply our results to various stochastic games with heterogeneous interactions, including systemic risk models with delays and stochastic network games., Comment: 58 pages

Large Language Models (LLMs) often struggle when prompted to generate content under specific constraints. However, in such cases it is often easy to check whether these constraints are satisfied or violated. Recent works have shown that LLMs can benefit from such "corrective feedback". Here we claim that this skill of LLMs can be significantly enhanced via training. We introduce an RL framework for teaching models to use such rewards, by simulating interaction sessions, and rewarding the model according to its ability to satisfy the constraints. We refer to our method as CORGI (Controlled Generation with RL for Guided Interaction), and evaluate it on a variety of controlled generation tasks using unlabeled training data. We find that CORGI consistently outperforms the baseline reinforcement learning method that does not incorporate conversational feedback. Furthermore, CORGI's interactive framework enables meta-learning, allowing the LLM to generalize better to guided interaction in new tasks. Our results clearly show that conversational optimization, when combined with reinforcement learning, significantly improves the effectiveness of LLMs in controlled generation contexts.

We study and classify algebraic families of Harish-Chandra pairs over the complex affine line and over the complex projective line with generic fiber that is isomorphic to the Harish-Chandra pair of $SL_2(\mathbb{R})$.

Face recognition technologies are increasingly used in various applications, yet they are vulnerable to face spoofing attacks. These spoofing attacks often involve unique 3D structures, such as printed papers or mobile device screens. Although stereo-depth cameras can detect such attacks effectively, their high-cost limits their widespread adoption. Conversely, two-sensor systems without extrinsic calibration offer a cost-effective alternative but are unable to calculate depth using stereo techniques. In this work, we propose a method to overcome this challenge by leveraging facial attributes to derive disparity information and estimate relative depth for anti-spoofing purposes, using non-calibrated systems. We introduce a multi-modal anti-spoofing model, coined Disparity Model, that incorporates created disparity maps as a third modality alongside the two original sensor modalities. We demonstrate the effectiveness of the Disparity Model in countering various spoof attacks using a comprehensive dataset collected from the Intel RealSense ID Solution F455. Our method outperformed existing methods in the literature, achieving an Equal Error Rate (EER) of 1.71% and a False Negative Rate (FNR) of 2.77% at a False Positive Rate (FPR) of 1%. These errors are lower by 2.45% and 7.94% than the errors of the best comparison method, respectively. Additionally, we introduce a model ensemble that addresses 3D spoof attacks as well, achieving an EER of 2.04% and an FNR of 3.83% at an FPR of 1%. Overall, our work provides a state-of-the-art solution for the challenging task of anti-spoofing in non-calibrated systems that lack depth information.

Intercalation is a robust method for tuning the physical properties of a vast number of van der Waals (vdW) materials. However, the prospects of using intercalation to modify magnetism in vdWs systems and the associated mechanisms have not been investigated adequately. In this work, we modulate magnetic order in an XY antiferromagnet NiPS3 single crystals by introducing pyridine molecules into the vdWs gap under different thermal conditions. X-ray diffraction measurements indicated pronounced changes in the lattice parameter beta, while magnetization measurements at in-plane and out-of-plane configurations exposed reversal trends in the crystals Neel temperatures through intercalation-de-intercalation processes. The changes in magnetic ordering were also supported by three-dimensional thermal diffusivity experiments. The preferred orientation of the pyridine dipoles within vdW gaps was deciphered via polarized Raman spectroscopy. The results highlight the relation between the preferential alignment of the intercalants, thermal transport, and crystallographic disorder along with the modulation of anisotropy in the magnetic order. The theoretical concept of double-exchange interaction in NiPS3 was employed to explain the intercalation-induced magnetic ordering. The study uncovers the merit of intercalation as a foundation for spin switches and spin transistors in advanced quantum devices.

The Quantum-Enhanced Space-Time (QUEST) experiment consists of a pair of co-located Power Recycled Michelson Interferometers, each designed to have a broadband, shot-noise limited displacement sensitivity of $2\times10^{-19}$ $\mathrm{m/\sqrt{Hz}}$ from 1 to 200 MHz. Here we present the first results of QUEST that set new upper limits on correlated length fluctuations from 13 to 80 MHz, constituting the first broadband constraints for a stochastic gravitational wave background at these frequencies. In a coincident observing run of $10^{4}$ s the averaging of the cross-correlation spectra between the two interferometer signals resulted in a strain sensitivity of $3\times10^{-20}$ $\mathrm{1/\sqrt{Hz}}$ at 40 MHz, making QUEST the most sensitive table-top interferometric system to date.

This paper contains a survey of supersingular isogeny graphs associated to supersingular elliptic curves and their various applications to cryptography. Within limitation of space, we attempt to address a broad audience and make this part widely accessible. For those graphs we also present three recent results and sketch their proofs. We then discuss a generalization to superspecial isogeny graphs associated to superspecial abelian varieties with real multiplication. These graphs were introduced by Charles, Goren and Lauter and so our discussion is brief. Motivated by their cryptographic applications, we prove a general theorem concerning generation of lattices over totally real fields by elements of specified norm. Throughout the paper we have attempted to clarify certain considerations that are either vaguely stated in the literature, or are folklore. We hope this paper will be useful both to a novice wishing to familiarize themselves with this very active area, and to the expert who may enjoy some vignettes and an overview of some new results., Comment: 62 pages, 5 figures. Submitted to the proceedings of the Advanced Research Workshop: Isogeny based post-quantum cryptography (July 2024)

Wildfires pose a significant natural disaster risk to populations and contribute to accelerated climate change. As wildfires are also affected by climate change, extreme wildfires are becoming increasingly frequent. Although they occur less frequently globally than those sparked by human activities, lightning-ignited wildfires play a substantial role in carbon emissions and account for the majority of burned areas in certain regions. While existing computational models, especially those based on machine learning, aim to predict lightning-ignited wildfires, they are typically tailored to specific regions with unique characteristics, limiting their global applicability. In this study, we present machine learning models designed to characterize and predict lightning-ignited wildfires on a global scale. Our approach involves classifying lightning-ignited versus anthropogenic wildfires, and estimating with high accuracy the probability of lightning to ignite a fire based on a wide spectrum of factors such as meteorological conditions and vegetation. Utilizing these models, we analyze seasonal and spatial trends in lightning-ignited wildfires shedding light on the impact of climate change on this phenomenon. We analyze the influence of various features on the models using eXplainable Artificial Intelligence (XAI) frameworks. Our findings highlight significant global differences between anthropogenic and lightning-ignited wildfires. Moreover, we demonstrate that, even over a short time span of less than a decade, climate changes have steadily increased the global risk of lightning-ignited wildfires. This distinction underscores the imperative need for dedicated predictive models and fire weather indices tailored specifically to each type of wildfire.

The $(2+1)$D Solid-On-Solid (SOS) model famously exhibits a roughening transition: on an $N\times N$ torus with the height at the origin rooted at $0$, the variance of $h(x)$, the height at $x$, is $O(1)$ at large inverse-temperature $\beta$, vs. $\asymp \log |x|$ at small $\beta$ (as in the Gaussian free field (GFF)). The former--rigidity at large $\beta$--is known for a wide class of $|\nabla\phi|^p$ models ($p=1$ being SOS) yet is believed to fail once the surface is on a slope (tilted boundary conditions). It is conjectured that the slope would destabilize the rigidity and induce the GFF-type behavior of the surface at small $\beta$. The only rigorous result on this is by Sheffield (2005): for these models of integer height functions, if the slope $\theta$ is irrational, then Var$(h(x))\to\infty$ with $|x|$ (with no known quantitative bound). We study a family of SOS surfaces at a large enough fixed $\beta$, on an $N\times N$ torus with a nonzero boundary condition slope $\theta$, perturbed by a potential $V$ of strength $\epsilon_\beta$ per site (arbitrarily small). Our main result is (a) the measure on the height gradients $\nabla h$ has a weak limit $\mu_\infty$ as $N\to\infty$; and (b) the scaling limit of a sample from $\mu_\infty$ converges to a full plane GFF. In particular, we recover the asymptotics Var$(h(x))\sim c\log|x|$. To our knowledge, this is the first example of a tilted $|\nabla\phi|^p$ model, or a perturbation thereof, where the limit is recovered at large $\beta$. The proof looks at random monotone surfaces that approximate the SOS surface, and shows that (i) these form a weakly interacting dimer model, and (ii) the renormalization framework of Giuliani, Mastropietro and Toninelli (2017) leads to the GFF limit. New ingredients are needed in both parts, including a nontrivial extension of [GMT17] from finite interactions to any long range summable interactions., Comment: 95 pages; 18 figures

We study the entropic repulsion of the low temperature 3D Ising and Potts interface in an $n\times n \times n$ box with blue boundary conditions on its bottom face (the hard floor), and red boundary conditions on its other five faces. For Ising, Frohlich and Pfister proved in 1987 that the typical interface height above the origin diverges (non-quantitatively), via correlation inequalities special to the Ising model; no such result was known for Potts. We show for both the Ising and Potts models that the entropic repulsion fully overcomes the potentially attractive interaction with the floor, and obtain a logarithmically diverging lower bound on the typical interface height. This is complemented by a conjecturally sharp upper bound of $\lfloor \xi^{-1}\log n\rfloor$ where $\xi$ is the rate function for a point-to-plane non-red connection under the infinite volume red measure. The proof goes through a coupled random-cluster interface to overcome the potential attractive interaction with the boundary, and a coupled fuzzy Potts model to reduce the upper bound to a simpler setting where the repulsion is attained by conditioning a no-floor interface to lie in the upper half-space., Comment: 37 pages; 1 figure

As artificial intelligence (AI) continues to rapidly advance, there is a growing demand to integrate AI capabilities into existing business applications. However, a significant gap exists between the rapid progress in AI and how slowly AI is being embedded into business environments. Deploying well-performing lab models into production settings, especially in on-premise environments, often entails specialized expertise and imposes a heavy burden of model management, creating significant barriers to implementing AI models in real-world applications. KModels leverages proven libraries and platforms (Kubeflow Pipelines, KServe) to streamline AI adoption by supporting both AI developers and consumers. It allows model developers to focus solely on model development and share models as transportable units (Templates), abstracting away complex production deployment concerns. KModels enables AI consumers to eliminate the need for a dedicated data scientist, as the templates encapsulate most data science considerations while providing business-oriented control. This paper presents the architecture of KModels and the key decisions that shape it. We outline KModels' main components as well as its interfaces. Furthermore, we explain how KModels is highly suited for on-premise deployment but can also be used in cloud environments. The efficacy of KModels is demonstrated through the successful deployment of three AI models within an existing Work Order Management system. These models operate in a client's data center and are trained on local data, without data scientist intervention. One model improved the accuracy of Failure Code specification for work orders from 46% to 83%, showcasing the substantial benefit of accessible and localized AI solutions.

T1 mapping is a valuable quantitative MRI technique for diagnosing diffuse myocardial diseases. Traditional methods, relying on breath-hold sequences and echo triggering, face challenges with patient compliance and arrhythmias, limiting their effectiveness. Image registration can enable motion-robust T1 mapping, but inherent intensity differences between time points pose a challenge. We introduce MBSS-T1, a self-supervised model for motion correction in cardiac T1 mapping, constrained by physical and anatomical principles. The physical constraints ensure expected signal decay behavior, while the anatomical constraints maintain realistic deformations. The unique combination of these constraints ensures accurate T1 mapping along the longitudinal relaxation axis. MBSS-T1 outperformed baseline deep-learning-based image registration approaches in a 5-fold experiment on a public dataset of 210 patients (STONE sequence) and an internal dataset of 19 patients (MOLLI sequence). MBSS-T1 excelled in model fitting quality ($R^2$: 0.975 vs. 0.941, 0.946), anatomical alignment (Dice score: 0.89 vs. 0.84, 0.88), and expert visual quality assessment for the presence of visible motion artifacts (4.33 vs. 3.38, 3.66). MBSS-T1 has the potential to enable motion-robust T1 mapping for a broader range of patients, overcoming challenges such as arrhythmias and suboptimal compliance, and allowing for free-breathing T1 mapping without requiring large training datasets. Our code will be publicly available upon acceptance.

There is a growing line of research on verifying the correctness of language models' outputs. At the same time, LMs are being used to tackle complex queries that require reasoning. We introduce CoverBench, a challenging benchmark focused on verifying LM outputs in complex reasoning settings. Datasets that can be used for this purpose are often designed for other complex reasoning tasks (e.g., QA) targeting specific use-cases (e.g., financial tables), requiring transformations, negative sampling and selection of hard examples to collect such a benchmark. CoverBench provides a diversified evaluation for complex claim verification in a variety of domains, types of reasoning, relatively long inputs, and a variety of standardizations, such as multiple representations for tables where available, and a consistent schema. We manually vet the data for quality to ensure low levels of label noise. Finally, we report a variety of competitive baseline results to show CoverBench is challenging and has very significant headroom. The data is available at https://huggingface.co/datasets/google/coverbench ., Comment: Huggingface Datasets link: https://huggingface.co/datasets/google/coverbench

We report the investigation of electronic collective modes in rare-earth-based magnets (Mo$_{2/3}$RE$_{1/3}$)$_2$AlC (also known as RE-$i$-MAX phases), where RE=Gd, Yb, and Dy, using single crystal samples. A detailed investigation of the Raman spectra of Gd-$i$-MAX samples at low temperatures, with a focus on the phonon behavior in relation to the antiferromagnetic (AFM) phase transition at 26 K is presented. Significant shifts in the central frequencies of several low-frequency phonon modes were observed below 25 K, correlating with the N\'{e}el transition. Integrated Raman intensity measurements indicated a reduction in the electronic background below the AFM transition temperature, suggesting the opening of a magnetic gap. Our analysis showed no new phonon modes. Therefore, we do not see any indication of a Brillouin zone folding of phonon mode to the $\Gamma$-point in our measurement. However, the hardening of all phonon modes at low temperatures points to a strong spin-phonon coupling effect. Using a temperature-dependent model of phonon frequency, we determined the spin-phonon coupling constant $\lambda$ to be less than 0.1 cm$^{-1}$ for all frequencies, which is of the same order of magnitude as found in other antiferromagnetic materials such as MnF$_{2}$ and FeF$_{2}$ with $T_N=68~K$ and $T_N=78~K$, respectively, but significantly lower than that of $CuO$ with $T_N=213~K$., Comment: 9 pages, 13 figures

A master equation containing a nonlinear term that gives rise to disentanglement has been recently investigated. In this study, a modified version, which is applicable for indistinguishable particles, is proposed, and explored for both the Bose-Hubbard and the Fermi-Hubbard models. It is found for both Bosons and Fermions that disentanglement can give rise to quantum phase transitions., Comment: arXiv admin note: text overlap with arXiv:2404.17580

Recent advancements in Language Models (LMs) have catalyzed the creation of multiple benchmarks, designed to assess these models' general capabilities. A crucial task, however, is assessing the validity of the benchmarks themselves. This is most commonly done via Benchmark Agreement Testing (BAT), where new benchmarks are validated against established ones using some agreement metric (e.g., rank correlation). Despite the crucial role of BAT for benchmark builders and consumers, there are no standardized procedures for such agreement testing. This deficiency can lead to invalid conclusions, fostering mistrust in benchmarks and upending the ability to properly choose the appropriate benchmark to use. By analyzing over 40 prominent benchmarks, we demonstrate how some overlooked methodological choices can significantly influence BAT results, potentially undermining the validity of conclusions. To address these inconsistencies, we propose a set of best practices for BAT and demonstrate how utilizing these methodologies greatly improves BAT robustness and validity. To foster adoption and facilitate future research,, we introduce BenchBench, a python package for BAT, and release the BenchBench-leaderboard, a meta-benchmark designed to evaluate benchmarks using their peers. Our findings underscore the necessity for standardized BAT, ensuring the robustness and validity of benchmark evaluations in the evolving landscape of language model research. BenchBench Package: github.com/IBM/BenchBench Leaderboard: hf.co/spaces/IBM/BenchBench, Comment: Under Review

We present HebDB, a weakly supervised dataset for spoken language processing in the Hebrew language. HebDB offers roughly 2500 hours of natural and spontaneous speech recordings in the Hebrew language, consisting of a large variety of speakers and topics. We provide raw recordings together with a pre-processed, weakly supervised, and filtered version. The goal of HebDB is to further enhance research and development of spoken language processing tools for the Hebrew language. Hence, we additionally provide two baseline systems for Automatic Speech Recognition (ASR): (i) a self-supervised model; and (ii) a fully supervised model. We present the performance of these two methods optimized on HebDB and compare them to current multi-lingual ASR alternatives. Results suggest the proposed method reaches better results than the evaluated baselines considering similar model sizes. Dataset, code, and models are publicly available under https://pages.cs.huji.ac.il/adiyoss-lab/HebDB/., Comment: Accepted at Interspeech2024

Maglaras, Moallemi, and Zheng (2021) have introduced a flexible queueing model for fragmented limit-order markets, whose fluid limit remains remarkably tractable. In the present study we prove that, in the limit of small and frequent orders, the discrete system indeed converges to the fluid limit, which is characterized by a system of coupled nonlinear ODEs with singular coefficients at the origin. Moreover, we establish that the fluid system is asymptotically stable for an arbitrary number of limit order books in that, over time, it converges to the stationary equilibrium state studied by Maglaras et al. (2021).

We consider a central trading desk which aggregates the inflow of clients' orders with unobserved toxicity, i.e. persistent adverse directionality. The desk chooses either to internalise the inflow or externalise it to the market in a cost effective manner. In this model, externalising the order flow creates both price impact costs and an additional market feedback reaction for the inflow of trades. The desk's objective is to maximise the daily trading P&L subject to end of the day inventory penalization. We formulate this setting as a partially observable stochastic control problem and solve it in two steps. First, we derive the filtered dynamics of the inventory and toxicity, projected to the observed filtration, which turns the stochastic control problem into a fully observed problem. Then we use a variational approach in order to derive the unique optimal trading strategy. We illustrate our results for various scenarios in which the desk is facing momentum and mean-reverting toxicity. Our implementation shows that the P&L performance gap between the partially observable problem and the full information case are of order $0.01\%$ in all tested scenarios., Comment: 52 pages, 10 figures

In the usual picture of Hawking radiation, the emission is spontaneous; it is caused by nothing. In contrast, the radiation from the ringdown after a black-hole merger is caused dynamically by the fluctuations of the event horizon. We explore the possibility that Hawking radiation is also emitted dynamically by horizon fluctuations, in the form of quasinormal modes. In fact, we find that the fundamental, least-damped quasinormal mode is sufficient to radiate the entire Hawking spectrum of photons and gravitons, since the quasinormal mode spectrum is broadened by damping. The resulting Hawking spectra are accurate with no graybody factors. By comparing Hawking radiation to ringdowns, we find that the quantum fluctuations of the horizon should be on the order of 0.1 Planck lengths. We compare this result with predictions ranging over 60 orders of magnitude. Further support for the model is provided by a sonic black hole experiment, in which additional horizon fluctuations are seen to produce Hawking-like radiation., Comment: 15 pages, 5 figures. This version discusses photons in addition to gravitons. More emphasis is placed on the Hawking spectra. The order of the sections is changed for clarity

Forced alignment (FA) plays a key role in speech research through the automatic time alignment of speech signals with corresponding text transcriptions. Despite the move towards end-to-end architectures for speech technology, FA is still dominantly achieved through a classic GMM-HMM acoustic model. This work directly compares alignment performance from leading automatic speech recognition (ASR) methods, WhisperX and Massively Multilingual Speech Recognition (MMS), against a Kaldi-based GMM-HMM system, the Montreal Forced Aligner (MFA). Performance was assessed on the manually aligned TIMIT and Buckeye datasets, with comparisons conducted only on words correctly recognized by WhisperX and MMS. The MFA outperformed both WhisperX and MMS, revealing a shortcoming of modern ASR systems. These findings highlight the need for advancements in forced alignment and emphasize the importance of integrating traditional expertise with modern innovation to foster progress. Index Terms: forced alignment, phoneme alignment, word alignment

Machine learning approaches to Structure-Based Drug Design (SBDD) have proven quite fertile over the last few years. In particular, diffusion-based approaches to SBDD have shown great promise. We present a technique which expands on this diffusion approach in two crucial ways. First, we address the size disparity between the drug molecule and the target/receptor, which makes learning more challenging and inference slower. We do so through the notion of a Virtual Receptor, which is a compressed version of the receptor; it is learned so as to preserve key aspects of the structural information of the original receptor, while respecting the relevant group equivariance. Second, we incorporate a protein language embedding used originally in the context of protein folding. We experimentally demonstrate the contributions of both the virtual receptors and the protein embeddings: in practice, they lead to both better performance, as well as significantly faster computations.

Chest X-rays or chest radiography (CXR), commonly used for medical diagnostics, typically enables limited imaging compared to computed tomography (CT) scans, which offer more detailed and accurate three-dimensional data, particularly contrast-enhanced scans like CT Pulmonary Angiography (CTPA). However, CT scans entail higher costs, greater radiation exposure, and are less accessible than CXRs. In this work we explore cross-modal translation from a 2D low contrast-resolution X-ray input to a 3D high contrast and spatial-resolution CTPA scan. Driven by recent advances in generative AI, we introduce a novel diffusion-based approach to this task. We evaluate the models performance using both quantitative metrics and qualitative feedback from radiologists, ensuring diagnostic relevance of the generated images. Furthermore, we employ the synthesized 3D images in a classification framework and show improved AUC in a PE categorization task, using the initial CXR input. The proposed method is generalizable and capable of performing additional cross-modality translations in medical imaging. It may pave the way for more accessible and cost-effective advanced diagnostic tools. The code for this project is available: https://github.com/NoaCahan/X-ray2CTPA ., Comment: preprint, project code: https://github.com/NoaCahan/X-ray2CTPA

Fine-grained visual classification (FGVC) involves classifying closely related sub-classes. This task is difficult due to the subtle differences between classes and the high intra-class variance. Moreover, FGVC datasets are typically small and challenging to gather, thus highlighting a significant need for effective data augmentation. Recent advancements in text-to-image diffusion models offer new possibilities for augmenting classification datasets. While these models have been used to generate training data for classification tasks, their effectiveness in full-dataset training of FGVC models remains under-explored. Recent techniques that rely on Text2Image generation or Img2Img methods, often struggle to generate images that accurately represent the class while modifying them to a degree that significantly increases the dataset's diversity. To address these challenges, we present SaSPA: Structure and Subject Preserving Augmentation. Contrary to recent methods, our method does not use real images as guidance, thereby increasing generation flexibility and promoting greater diversity. To ensure accurate class representation, we employ conditioning mechanisms, specifically by conditioning on image edges and subject representation. We conduct extensive experiments and benchmark SaSPA against both traditional and recent generative data augmentation methods. SaSPA consistently outperforms all established baselines across multiple settings, including full dataset training, contextual bias, and few-shot classification. Additionally, our results reveal interesting patterns in using synthetic data for FGVC models; for instance, we find a relationship between the amount of real data used and the optimal proportion of synthetic data. Code is available at https://github.com/EyalMichaeli/SaSPA-Aug., Comment: Under review. Code is available at https://github.com/EyalMichaeli/SaSPA-Aug

In this note, we study the low temperature $(2+1)$D SOS interface above a hard floor with critical pinning potential $\lambda_w= \log (\frac{1}{1-e^{-4\beta}})$. At $\lambda<\lambda_w$ entropic repulsion causes the surface to delocalize and be rigid at height $\frac1{4\beta}\log n+O(1)$; at $\lambda>\lambda_w$ it is localized at some $O(1)$ height. We show that at $\lambda=\lambda_w$, there is delocalization, with rigidity now at height $\lfloor \frac1{6\beta}\log n+\frac13\rfloor$, confirming a conjecture of Lacoin., Comment: 12 pages

Blockchains implement monetary systems operated by committees of nodes. The robustness of established blockchains presents an opportunity to leverage their infrastructure for creating expansion chains. Expansion chains can provide additional functionality to the primary chain they leverage or implement separate functionalities, while benefiting from the primary chain's security and the stability of its tokens. Indeed, tools like Ethereum's EigenLayer enable nodes to stake (deposit collateral) on a primary chain to form a committee responsible for operating an expansion chain. But here is the rub. Classical protocols assume correct, well-behaved nodes stay correct indefinitely. Yet in our case, the stake incentivizes correctness--it will be slashed (revoked) if its owner deviates. Once a node withdraws its stake, there is no basis to assume its correctness. To address the new challenge, we present Aegis, an expansion chain based on primary-chain stake, assuming a bounded primary-chain write time. Aegis uses references from Aegis blocks to primary blocks to define committees, checkpoints on the primary chain to perpetuate decisions, and resets on the primary chain to establish a new committee if the previous one becomes obsolete. It ensures safety at all times and rapid progress when latency among Aegis nodes is low.

Automated Market Makers (AMMs) are a cornerstone of decentralized finance. They are smart contracts (stateful programs) running on blockchains. They enable virtual token exchange: Traders swap tokens with the AMM for a fee, while liquidity providers supply liquidity and earn these fees. Demand for AMMs is growing rapidly, but our experiment-based estimates show that current architectures cannot meet the projected demand by 2029. This is because the execution of existing AMMs is non-parallelizable. We present SAMM, an AMM comprising multiple shards. All shards are AMMs running on the same chain, but their independence enables parallel execution. Unlike classical sharding solutions, here security relies on incentive compatibility. Therefore, SAMM introduces a novel fee design. Through analysis of Subgame-Perfect Nash Equilibria (SPNE), we show that SAMM incentivizes the desired behavior: Liquidity providers balance liquidity among all shards, overcoming destabilization attacks, and trades are evenly distributed. We validate our game-theoretic analysis with a simulation using real-world data. We evaluate SAMM by implementing and deploying it on local testnets of the Sui and Solana blockchains. To our knowledge, this is the first quantification of ``hot-contract'' performance. SAMM improves throughput by 5x and 16x, respectively, potentially more with better parallelization of the underlying blockchains. It is directly deployable, mitigating the upcoming scaling bottleneck.

Large Language Models (LLMs) often do not perform well on queries that require the aggregation of information across texts. To better evaluate this setting and facilitate modeling efforts, we introduce TACT - Text And Calculations through Tables, a dataset crafted to evaluate LLMs' reasoning and computational abilities using complex instructions. TACT contains challenging instructions that demand stitching information scattered across one or more texts, and performing complex integration on this information to generate the answer. We construct this dataset by leveraging an existing dataset of texts and their associated tables. For each such tables, we formulate new queries, and gather their respective answers. We demonstrate that all contemporary LLMs perform poorly on this dataset, achieving an accuracy below 38%. To pinpoint the difficulties and thoroughly dissect the problem, we analyze model performance across three components: table-generation, Pandas command-generation, and execution. Unexpectedly, we discover that each component presents substantial challenges for current LLMs. These insights lead us to propose a focused modeling framework, which we refer to as IE as a tool. Specifically, we propose to add "tools" for each of the above steps, and implement each such tool with few-shot prompting. This approach shows an improvement over existing prompting techniques, offering a promising direction for enhancing model capabilities in these tasks., Comment: Accepted to NeurIPS 2024. Website (https://tact-benchmark.github.io), Huggingface (https://huggingface.co/datasets/google/TACT)

In this paper, we propose an efficient, fast, and versatile distillation method to accelerate the generation of pre-trained diffusion models: Flash Diffusion. The method reaches state-of-the-art performances in terms of FID and CLIP-Score for few steps image generation on the COCO2014 and COCO2017 datasets, while requiring only several GPU hours of training and fewer trainable parameters than existing methods. In addition to its efficiency, the versatility of the method is also exposed across several tasks such as text-to-image, inpainting, face-swapping, super-resolution and using different backbones such as UNet-based denoisers (SD1.5, SDXL) or DiT (Pixart-$\alpha$), as well as adapters. In all cases, the method allowed to reduce drastically the number of sampling steps while maintaining very high-quality image generation. The official implementation is available at https://github.com/gojasper/flash-diffusion., Comment: Accepted to AAAI 2025

A long-standing open conjecture of Branko Gr\"unbaum from 1972 states that any simple arrangement of $n$ pairwise intersecting pseudocircles in the plane can have at most $2n-2$ digons. Agarwal et al. proved this conjecture for arrangements of pairwise intersecting pseudocircles in which there is a common point surrounded by all pseudocircles. Recently, Felsner, Roch and Scheucher showed that Gr\"unbaum's conjecture is true for arrangements of pairwise intersecting pseudocircles in which there are three pseudocircles every pair of which create a digon. In this paper we prove this over 50-year-old conjecture of Gr\"unbaum for any simple arrangement of pairwise intersecting circles in the plane.

A central problem in quantum mechanics involves solving the Electronic Schrodinger Equation for a molecule or material. The Variational Monte Carlo approach to this problem approximates a particular variational objective via sampling, and then optimizes this approximated objective over a chosen parameterized family of wavefunctions, known as the ansatz. Recently neural networks have been used as the ansatz, with accompanying success. However, sampling from such wavefunctions has required the use of a Markov Chain Monte Carlo approach, which is inherently inefficient. In this work, we propose a solution to this problem via an ansatz which is cheap to sample from, yet satisfies the requisite quantum mechanical properties. We prove that a normalizing flow using the following two essential ingredients satisfies our requirements: (a) a base distribution which is constructed from Determinantal Point Processes; (b) flow layers which are equivariant to a particular subgroup of the permutation group. We then show how to construct both continuous and discrete normalizing flows which satisfy the requisite equivariance. We further demonstrate the manner in which the non-smooth nature ("cusps") of the wavefunction may be captured, and how the framework may be generalized to provide induction across multiple molecules. The resulting theoretical framework entails an efficient approach to solving the Electronic Schrodinger Equation., Comment: Added references

Multistability cannot be derived from any theoretical model that is based on a monostable master equation. On the other hand, multistability is experimentally-observed in a variety of quantum systems. A master equation having a nonlinear term that gives rise to disentanglement has been recently proposed . The dynamics governed by this master equation is explored for a quantum system made of coupled spins. It is found that the added nonlinear term can give rise to multistability. The spins' response to an externally applied magnetic field is evaluated, and both a phase transition and a dynamical instability are found. These findings, which originate from disentanglement-induced multistability, indirectly support the hypothesis that spontaneous disentanglement occurs in quantum systems.

Dense vector representations for sentences made significant progress in recent years as can be seen on sentence similarity tasks. Real-world phrase retrieval applications, on the other hand, still encounter challenges for effective use of dense representations. We show that when target phrases reside inside noisy context, representing the full sentence with a single dense vector, is not sufficient for effective phrase retrieval. We therefore look into the notion of representing multiple, sub-sentence, consecutive word spans, each with its own dense vector. We show that this technique is much more effective for phrase mining, yet requires considerable compute to obtain useful span representations. Accordingly, we make an argument for contextualized word/token embeddings that can be aggregated for arbitrary word spans while maintaining the span's semantic meaning. We introduce a modification to the common contrastive loss used for sentence embeddings that encourages word embeddings to have this property. To demonstrate the effect of this method we present a dataset based on the STS-B dataset with additional generated text, that requires finding the best matching paraphrase residing in a larger context and report the degree of similarity to the origin phrase. We demonstrate on this dataset, how our proposed method can achieve better results without significant increase to compute.

When large language models are aligned via supervised fine-tuning, they may encounter new factual information that was not acquired through pre-training. It is often conjectured that this can teach the model the behavior of hallucinating factually incorrect responses, as the model is trained to generate facts that are not grounded in its pre-existing knowledge. In this work, we study the impact of such exposure to new knowledge on the capability of the fine-tuned model to utilize its pre-existing knowledge. To this end, we design a controlled setup, focused on closed-book QA, where we vary the proportion of the fine-tuning examples that introduce new knowledge. We demonstrate that large language models struggle to acquire new factual knowledge through fine-tuning, as fine-tuning examples that introduce new knowledge are learned significantly slower than those consistent with the model's knowledge. However, we also find that as the examples with new knowledge are eventually learned, they linearly increase the model's tendency to hallucinate. Taken together, our results highlight the risk in introducing new factual knowledge through fine-tuning, and support the view that large language models mostly acquire factual knowledge through pre-training, whereas fine-tuning teaches them to use it more efficiently., Comment: Accepted as a long paper at EMNLP 2024

A coherent optical spectrum analyzer is integrated with a rotating quarter wave plate polarimeter. The combined polarimeter optical spectrum analyzer (POSA) allows the extraction of the state of polarization with high spectral resolution. POSA is used in this work to study two optical systems. The first is an optical modulator based on a ferrimagnetic sphere resonator. POSA is employed to explore the underlying magneto-optical mechanism responsible for modulation sideband asymmetry. The second system under study is a cryogenic fiber loop laser, which produces an unequally spaced optical comb. Polarization measurements provide insights on the nonlinear processes responsible for comb creation. Characterizations extracted from POSA data provide guidelines for performance optimization of applications based on these systems under study.

This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. We discuss our approach toward post-training and deploying Gemini models responsibly to users through services including Gemini, Gemini Advanced, Google AI Studio, and Cloud Vertex AI.

Whisper is a state-of-the-art automatic speech recognition (ASR) model (Radford et al., 2022). Although Swiss German dialects are allegedly not part of Whisper's training data, preliminary experiments showed that Whisper can transcribe Swiss German quite well, with the output being a speech translation into Standard German. To gain a better understanding of Whisper's performance on Swiss German, we systematically evaluate it using automatic, qualitative, and human evaluation. We test its performance on three existing test sets: SwissDial (Dogan-Sch\"onberger et al., 2021), STT4SG-350 (Pl\"uss et al., 2023), and Swiss Parliaments Corpus (Pl\"uss et al., 2021). In addition, we create a new test set for this work, based on short mock clinical interviews. For automatic evaluation, we used word error rate (WER) and BLEU. In the qualitative analysis, we discuss Whisper's strengths and weaknesses and anylyze some output examples. For the human evaluation, we conducted a survey with 28 participants who were asked to evaluate Whisper's performance. All of our evaluations suggest that Whisper is a viable ASR system for Swiss German, so long as the Standard German output is desired., Comment: Accepted at VarDial 2024 (the eleventh Workshop on NLP for Similar Languages, Varieties and Dialects 2024), Mexico City

Background: Recent advancements in Artificial Intelligence (AI) contributed significantly to suicide assessment, however, our theoretical understanding of this complex behavior is still limited. Objective: This study aimed to harness AI methodologies to uncover hidden risk factors that trigger or aggravate suicide behaviors. Method: The primary dataset included 228,052 Facebook postings by 1,006 users who completed the gold-standard Columbia Suicide Severity Rating Scale. This dataset was analyzed using a bottom-up research pipeline without a-priory hypotheses and its findings were validated using a top-down analysis of a new dataset. This secondary dataset included responses by 1,062 participants to the same suicide scale as well as to well-validated scales measuring depression and boredom. Results: An almost fully automated, AI-guided research pipeline resulted in four Facebook topics that predicted the risk of suicide, of which the strongest predictor was boredom. A comprehensive literature review using APA PsycInfo revealed that boredom is rarely perceived as a unique risk factor of suicide. A complementing top-down path analysis of the secondary dataset uncovered an indirect relationship between boredom and suicide, which was mediated by depression. An equivalent mediated relationship was observed in the primary Facebook dataset as well. However, here, a direct relationship between boredom and suicide risk was also observed. Conclusions: Integrating AI methods allowed the discovery of an under-researched risk factor of suicide. The study signals boredom as a maladaptive 'ingredient' that might trigger suicide behaviors, regardless of depression. Further studies are recommended to direct clinicians' attention to this burdening, and sometimes existential experience.

Precision measurements of space and time, like those made by the detectors of the Laser Interferometer Gravitational-wave Observatory (LIGO), are often confronted with fundamental limitations imposed by quantum mechanics. The Heisenberg uncertainty principle dictates that the position and momentum of an object cannot both be precisely measured, giving rise to an apparent limitation called the Standard Quantum Limit (SQL). Reducing quantum noise below the SQL in gravitational-wave detectors, where photons are used to continuously measure the positions of freely falling mirrors, has been an active area of research for decades. Here we show how the LIGO A+ upgrade reduced the detectors' quantum noise below the SQL by up to 3 dB while achieving a broadband sensitivity improvement, more than two decades after this possibility was first presented.

One mechanistic interpretation of baroclinic instability is that of mutual constructive interference of Rossby edge-waves. While the two edge-waves and their relative phase-shifts are invoked as part of the mechanistic interpretation, the phase-tilts of the related normal modes are often presented instead. Here we highlight the differences between edge-wave phase-shifts and normal-mode phase-tilts, in the context of an Eady problem modified by the presence of a sloping boundary. We argue and present evidence that the normal-mode phase-tilt is potentially a misleading quantity to use, and edge-wave phase-shifts should be the ones that are mechanistically relevant. We also provide a clarification for the mechanistic rationalization for baroclinic instability in the presence of slopes (such as suppression of growth rates) that is valid over all parameter space, in contrast to previous attempts. We further present evidence that there is a strong correlation between quantities diagnosed from the GEOMETRIC framework with the edge-wave phase-shifts, but not the normal-mode phase-tilts. The result is noteworthy in that the geometric framework makes no explicit reference to the edge-wave structures in its construction, but the correlation suggests that in problems where edge-wave structures are not so well-defined or readily available, the GEOMETRIC framework should still capture mechanistic and dynamical information. Some implications for parameterization of baroclinic instability and relevant eddy-mean feedbacks are discussed. For completeness, we also provide an explicit demonstration that the linear instability problem of the present modified Eady problem is parity-time symmetric, and speculate on some suggestive links between parity-time symmetry, shear instability, and the edge-wave interaction mechanism., Comment: 9 figures; discrepancy of arXiv abstract and file abstract, because number of characters exceeded in arXiv version