Bar-Ilan Faculty of Engineering

Temporal alignment of sequences is a fundamental challenge in many applications, such as computer vision and bioinformatics, where local time shifting needs to be accounted for. Misalignment can lead to poor model generalization, especially in high-dimensional sequences. Existing methods often struggle with optimization when dealing with high-dimensional sparse data, falling into poor alignments. Feature selection is frequently used to enhance model performance for sparse data. However, a fixed set of selected features would not generally work for dynamically changing sequences and would need to be modified based on the state of the sequence. Therefore, modifying the selected feature based on contextual input would result in better alignment. Our suggested method, Conditional Deep Canonical Temporal Time Warping (CDCTW), is designed for temporal alignment in sparse temporal data to address these challenges. CDCTW enhances alignment accuracy for high dimensional time-dependent views be performing dynamic time warping on data embedded in maximally correlated subspace which handles sparsity with novel feature selection method. We validate the effectiveness of CDCTW through extensive experiments on various datasets, demonstrating superior performance over previous techniques.

Training and fine-tuning large language models (LLMs) come with challenges related to memory and computational requirements due to the increasing size of the model weights and the optimizer states. Various techniques have been developed to tackle these challenges, such as low-rank adaptation (LoRA), which involves introducing a parallel trainable low-rank matrix to the fixed pre-trained weights at each layer. However, these methods often fall short compared to the full-rank weight training approach, as they restrict the parameter search to a low-rank subspace. This limitation can disrupt training dynamics and require a full-rank warm start to mitigate the impact. In this paper, we introduce a new method inspired by a phenomenon we formally prove: as training progresses, the rank of the estimated layer gradients gradually decreases, and asymptotically approaches rank one. Leveraging this, our approach involves adaptively reducing the rank of the gradients during Adam optimization steps, using an efficient online-updating low-rank projections rule. We further present a randomized SVD scheme for efficiently finding the projection matrix. Our technique enables full-parameter fine-tuning with adaptive low-rank gradient updates, significantly reducing overall memory requirements during training compared to state-of-the-art methods while improving model performance in both pretraining and fine-tuning. Finally, we provide a convergence analysis of our method and demonstrate its merits for training and fine-tuning language and biological foundation models.

We study the question of whether a sequence of positive integers is the degree sequence of some outerplanar (a.k.a. 1-page book embeddable) graph G. If so, G is an outerplanar realization of d and d is an outerplanaric sequence. The case where is easy, as d has a realization by a forest (which is trivially an outerplanar graph). In this paper, we consider the family of all sequences d of even sum , where is the number of x’s in d. (The second inequality is a necessary condition for a sequence d with to be outerplanaric.) We partition into two disjoint subfamilies, , such that every sequence in is provably non-outerplanaric, and every sequence in is given a realizing graph G enjoying a 2-page book embedding (and moreover, one of the pages is also bipartite).

Fusing information from different modalities can enhance data analysis tasks, including clustering. However, existing multi-view clustering (MVC) solutions are limited to specific domains or rely on a suboptimal and computationally demanding two-stage procedure of representation and clustering. We propose an end-to-end deep learning-based MVC framework for general data (image, tabular, etc.). Our approach involves learning meaningful fused data representations with a novel permutation-based canonical correlation objective. Concurrently, we learn cluster assignments by identifying consistent pseudo-labels across multiple views. We demonstrate the effectiveness of our model using ten MVC benchmark datasets. Theoretically, we show that our model approximates the supervised linear discrimination analysis (LDA) representation. Additionally, we provide an error bound induced by false-pseudo label annotations.

The time-multiplexing super-resolution concept requires post-processing for extracting the super-resolved image. Moreover, to perform the post-processing image restoration, one needs to know the exact high-resolution encoding pattern. Both of these limiting conditions are overcome by the method and experiment reported in this letter.

An erratum to correct a mistake in the caption of Fig. 5. [Opt. Express 32, 12040 (2024)] 10.1364/OE.515419. The corrections have no influence on the results and conclusions of the original paper.

Source-free domain adaptation (SFDA) aims to adapt a source-trained model to an unlabeled target domain without access to the source data. SFDA has attracted growing attention in recent years, where existing approaches focus on self-training that usually includes pseudo-labeling techniques. In this paper, we introduce a novel noise-learning approach tailored to address noise distribution in domain adaptation settings and learn to de-confuse the pseudo-labels. More specifically, we learn a noise transition matrix of the pseudo-labels to capture the label corruption of each class and learn the underlying true label distribution. Estimating the noise transition matrix enables a better true class-posterior estimation, resulting in better prediction accuracy. We demonstrate the effectiveness of our approach when combined with several SFDA methods: SHOT, SHOT++, and AaD. We obtain state-of-the-art results on three domain adaptation datasets: VisDA, DomainNet, and OfficeHome.

Semi Insulating GaAs alpha detectors with anode GaAs P+ contact layer were fabricated and characterized. The contact layer growth was carried out by Metal Organic Chemical Vapor Deposition (MOCVD) and the detector performances were compared to the performances of a front Schottky contact detector. The front side Schottky contact suffers from electron injection into the GaAs substrate. This injection is eliminated by using a P+ anode blocking layer with an ohmic contact, resulting in a reduction of leakage current at reverse bias values of up to 70 V. For example, at 30 V the leakage currents were 50 nA/cm2 and 150 nA/cm2 for the ohmic and the Schottky anode detectors, respectively. For both detectors, the charge collection efficiency was increased by a factor of ~2 after grinding the substrates from 650 μm to 310 μm thickness, with no leakage current degradation. In addition, rapid thermal process (RTP …

We investigate enantiomers of chiral organo-metallic particles which exhibit collective memory effect. Under the influence of magnetic field, the millions of particles in solution form macroscopic shapes and when dispersed again at zero field they return to their original shape. The microparticles forming the shaped structures are collectively coupled under the influence of long-range van der Waals exchange interactions which govern the collective macroscopic structure. We recently have found that non-local quantum exchange interactions between particles persist up to ten meters, close to room temperature. Here, we investigate further the perception of time in the system. We find that the particles exhibit temporal correlations between the future and present states of the system. The forces which govern the collective memory effect and shape the macroscopic structure therefore allow to visualize quantum phenomena which extend the classical causality notion into an expanded nonlocal space-time reality. We propose that the observations are attributed to chirality-discriminating van der Waals exchange interactions coupled to vacuum fluctuations.

The inference of gene regulatory networks (GRNs) from single-cell RNAseq data allows for mechanistic characterization of the different cell states and their dynamics in complex biological processes. While numerous algorithms have been proposed to infer GRNs from single-cell transcriptomic data, multiple network solutions may explain the same dataset, posing a challenge for biologically meaningful interpretation. Here, we use the Reasoning Engine for Interaction Networks (RE:IN), a computational tool based on formal reasoning, to characterize GRN ensembles in the context of acute kidney injury (AKI). To this end, we applied RE:IN to a single-cell RNAseq dataset from a mouse ischemia reperfusion injury (IRI) model, focusing on distinct proximal tubule cell states related to kidney injury and repair. We first created an Abstract Boolean Network (ABN) model for the kidney using RE:IN and synthesized an …

The time-multiplexing super-resolution concept requires post-processing for extracting the super-resolved image. Moreover, to perform the post-processing image restoration, one needs to know the exact high-resolution encoding pattern. Both of these limiting conditions are overcome by the method and experiment reported in this letter.

The Ternary Content Addressable Memory (TCAM) is a crucial component of satellite communication systems. Space-oriented TCAMs face unique challenges, as they must operate within a very limited energy budget and are susceptible to high Soft Error Rates (SER) due to ionizing particle radiation. The Dual Interlocked Storage Cell (DICE) based memory is capable of withstanding soft errors. However, its reliability diminishes in presence of multiple node upsets. Moreover, recent studies indicate that DICE resilience to even single-node upsets degrades in advanced technology nodes. This issue is further exacerbated by the scaling of the supply voltage to reduce power consumption. In this paper, we propose SpaceCAM, a DICE-based TCAM that overcomes the above limitations and enables aggressive voltage scaling while withstanding multiple node upsets in each memory row. SpaceCAM enables soft error …

Interaction-free measurement (IFM) is a promising technique for low-dose detection and imaging, offering the unique advantage of probing an object with an overall reduced absorption of the interrogating photons. We propose an experiment to demonstrate IFM in the single x ray photon regime. The proposed scheme relies on the triple-Laue (LLL) symmetric x ray interferometer, where each Laue diffraction acts as a lossy beam splitter. In contrast to many quantum effects which are highly vulnerable to loss, we show that an experimental demonstration of this effect in the x ray regime is feasible and can achieve detection with reduced dose and high IFM efficiency even in the presence of substantial loss in the system. The latter aspect is claimed to be a general property of IFM based on our theoretical analysis. We scrutinize two suitable detection schemes that offer a dose reduction of up to half compared with direct …

Recognizing the need for innovative therapeutic approaches in the management of autoimmune diseases , our current investigation explores the potential of autologous extracellular vesicles (EVs), derived from blood of rheumatoid arthritis (RA) patients, to serve as therapeutic vectors to improve drug delivery. We found that circulating EVs derived from arthritic mice (Collagen-induced arthritis model) express the joint/synovia homing receptor, αVβ3 integrin. Importantly, both autologous labelled EVs, derived from blood of arthritic mice (Collagen antibody-induced arthritis model) and healthy mice-derived EVs, exhibit targeted migration toward inflamed synovia without infiltrating healthy joints, as demonstrated by an in-vivo imaging system. Furthermore, EVs derived from plasma of RA patients show an overexpression of αV integrin and are effectively taken up by LPS/TNFα-induced activated human synovial cell …

The time-multiplexing super-resolution concept requires post-processing for extracting the super-resolved image. Moreover, to perform the post-processing image restoration, one needs to know the exact high-resolution encoding pattern. Both of these limiting conditions are overcome by the method and experiment reported in this letter.

Epidemic spreading over populations networks has been an important subject of research for several decades, and especially during the Covid-19 pandemic. Most epidemic outbreaks are likely to create multiple mutations during their spreading over the population. In this paper, we study the evolution of a pathogen which can mutate continuously during the epidemic spreading. We consider pathogens whose mutating parameter is the mortality mean-time, and study the evolution of this parameter over the spreading process. We use analytical methods to compute the dynamic equation of the epidemic and the conditions for it to spread. We also use numerical simulations to study the pathogen flow in this case, and to understand the mutation phenomena. We show that the natural selection leads to less violent pathogens becoming predominant in the population. We discuss a wide range of network structures and show how different effects are manifested in each case.

Acknowledging that current analytical models alone cannot provide the performance and sophistication that state-of-the-art systems should be endowed with, in the last decade, we have witnessed a rapid paradigm shift from model-based algorithms to data-driven ones, using primarily deep neural networks (DNNs), with many successful solutions in diverse application areas, such as speech and audio enhancement, source separation and localization, dereverberation, sparse representations of audio signals, audio rendering, acoustic event detection, music information retrieval, and more. However, learning-based methods, especially those based on DNNs, do not usually embrace the physical nature of the problem and rather optimize the nonlinear relationship between training data and expected results, relying on only computational power. Many problems call for more efficient solutions that minimize the …

The Ternary Content Addressable Memory (TCAM) is a crucial component of satellite communication systems. Space-oriented TCAMs face unique challenges, as they must operate within a very limited energy budget and are susceptible to high Soft Error Rates (SER) due to ionizing particle radiation. The Dual Interlocked Storage Cell (DICE) based memory is capable of withstanding soft errors. However, its reliability diminishes in presence of multiple node upsets. Moreover, recent studies indicate that DICE resilience to even single-node upsets degrades in advanced technology nodes. This issue is further exacerbated by the scaling of the supply voltage to reduce power consumption. In this paper, we propose SpaceCAM, a DICE-based TCAM that overcomes the above limitations and enables aggressive voltage scaling while withstanding multiple node upsets in each memory row. SpaceCAM enables soft error …

Quantitative phase microscopy (QPM), a technique combining phase imaging and microscopy, enables visualization of the 3D topography in reflective samples, as well as the inner structure or refractive index distribution of transparent and translucent samples. Similar to other imaging modalities, QPM is constrained by the conflict between numerical aperture (NA) and field of view (FOV): an imaging system with a low NA has to be employed to maintain a large FOV. This fact severely limits the resolution in QPM up to being the illumination wavelength. Consequently, finer structures of samples cannot be resolved by using modest NA objectives in QPM. Aimed to that, many approaches, such as oblique illumination, structured illumination, and speckle illumination (just to cite a few), have been proposed to improve the spatial resolution (or the space bandwidth product) in phase microscopy by restricting other degrees of freedom (mostly time). This paper aims to provide an up to date review on the resolution enhancement approaches in QPM, discussing the pros and cons of each technique as well as the confusion on resolution definition claims on QPM and other coherent microscopy methods. Through this survey, we will review the most appealing and useful techniques for superresolution in coherent microscopy, working with and without lenses and with special attention to QPM.

The inference of gene regulatory networks (GRNs) from single-cell RNAseq data allows for mechanistic characterization of the different cell states and their dynamics in complex biological processes. While numerous algorithms have been proposed to infer GRNs from single-cell transcriptomic data, multiple network solutions may explain the same dataset, posing a challenge for biologically meaningful interpretation. Here, we use the Reasoning Engine for Interaction Networks (RE:IN), a computational tool based on formal reasoning, to characterize GRN ensembles in the context of acute kidney injury (AKI). To this end, we applied RE:IN to a single-cell RNAseq dataset from a mouse ischemia reperfusion injury (IRI) model, focusing on distinct proximal tubule cell states related to kidney injury and repair. We first created an Abstract Boolean Network (ABN) model for the kidney using RE:IN and synthesized an …

We address the problem of Reliable Broadcast in asynchronous message-passing systems with n nodes, of which up to t are malicious (faulty), in addition to a message adversary that can drop some of the messages sent by correct (non-faulty) nodes. We present a Message-Adversary-Tolerant Byzantine Reliable Broadcast (MBRB) algorithm that communicates O (| m|+ nκ) bits per node, where| m| represents the length of the application message and κ= Ω (log n) is a security parameter. This communication complexity is optimal up to the parameter κ. This significantly improves upon the state-of-the-art MBRB solution (Albouy, Frey, Raynal, and Taïani, TCS 2023), which incurs communication of O (n| m|+ n²κ) bits per node. Our solution sends at most 4n² messages overall, which is asymptotically optimal. Reduced communication is achieved by employing coding techniques that replace the need for all nodes to (re-) broadcast the entire application message m. Instead, nodes forward authenticated fragments of the encoding of m using an erasure-correcting code. Under the cryptographic assumptions of threshold signatures and vector commitments, and assuming n> 3t+ 2d, where the adversary drops at most d messages per broadcast, our algorithm allows at least 𝓁= n-t-(1+ ε) d (for any arbitrarily low ε> 0) correct nodes to reconstruct m, despite missing fragments caused by the malicious nodes and the message adversary.