Search

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

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

In the realm of automatic speech recognition (ASR), robustness in noisy environments remains a significant challenge. Recent ASR models, such as Whisper, have shown promise, but their efficacy in noisy conditions can be further enhanced. This study is focused on recovering from packet loss to improve the word error rate (WER) of ASR models. We propose using a front-end adaptation network connected to a frozen ASR model. The adaptation network is trained to modify the corrupted input spectrum by minimizing the criteria of the ASR model in addition to an enhancement loss function. Our experiments demonstrate that the adaptation network, trained on Whisper's criteria, notably reduces word error rates across domains and languages in packet-loss scenarios. This improvement is achieved with minimal affect to Whisper model's foundational performance, underscoring our method's practicality and potential in enhancing ASR models in challenging acoustic environments., Comment: Accepted for publication at INTERSPEECH 2024

Atomic interferometers measure phase differences along paths with exceptional precision. Tweezer interferometry represents a novel approach for this measurement by guiding particles along predefined trajectories. This study explores the feasibility of using condensed bosons in tweezer interferometry. Unlike the factor $\sqrt{N}$ enhancement expected with classical ensembles, using NOON state interferometry can yield an enhancement by a factor of $N$. We consider a protocol for a tweezer-based NOON state interferometer that includes adiabatic splitting and merging of condensed bosons, followed by adiabatic branching for phase encoding. Our theoretical analysis focuses on the conditions necessary to achieve adiabaticity and avoid spontaneous symmetry breaking. Additionally, we demonstrate the feasibility of the proposed scheme and estimate the time required to perform these sweep processes., Comment: 14 pages, 10 figures

Plant diseases are major causes of production losses and may have a significant impact on the agricultural sector. Detecting pests as early as possible can help increase crop yields and production efficiency. Several robotic monitoring systems have been developed allowing to collect data and provide a greater understanding of environmental processes. An agricultural robot can enable accurate timely detection of pests, by traversing the field autonomously and monitoring the entire cropped area within a field. However, in many cases it is impossible to sample all plants due to resource limitations. In this thesis, the development and evaluation of several sampling algorithms are presented to address the challenge of an agriculture-monitoring ground robot designed to locate insects in an agricultural field, where complete sampling of all the plants is infeasible. Two situations were investigated in simulation models that were specially developed as part of this thesis: where no a-priori information on the insects is available and where prior information on the insects distributions within the field is known. For the first situation, seven algorithms were tested, each utilizing an approach to sample the field without prior knowledge of it. For the second situation, we present the development and evaluation of a dynamic sampling algorithm which utilizes real-time information to prioritize sampling at suspected points, locate hot spots and adapt sampling plans accordingly. The algorithm's performance was compared to two existing algorithms using Tetranychidae insect data from previous research. Analyses revealed that the dynamic algorithm outperformed the others.

Quantum mechanically, a driving process is expected to be reversible in the quasistatic limit, also known as the adiabatic theorem. This statement stands in opposition to classical mechanics, where a mix of regular and chaotic dynamics implies irreversibility. A paradigm for demonstrating the signatures of chaos in quantum irreversibility is a sweep process whose objective is to transfer condensed bosons from a source orbital. We show that such a protocol is dominated by an interplay of adiabatic-shuttling and chaos-assisted depletion processes. The latter is implied by interaction terms that spoil the Bogoliubov integrability of the Hamiltonian. As the sweep rate is lowered, a crossover to a regime that is dominated by quantum fluctuations is encountered, featuring a breakdown of quantum-to-classical correspondence. The major aspects of this picture are not captured by the common two-orbital approximation, which implies failure of the familiar many-body Landau-Zener paradigm., Comment: 17 pages, 13 figures

Over the last few years, deep learning has grown in popularity for speaker verification, identification, and diarization. Inarguably, a significant part of this success is due to the demonstrated effectiveness of their speaker representations. These, however, are heavily dependent on large amounts of annotated data and can be sensitive to new domains. This study proposes an entirely unsupervised deep-learning model for speaker diarization. Specifically, the study focuses on generating high-quality neural speaker representations without any annotated data, as well as on estimating secondary hyperparameters of the model without annotations. The speaker embeddings are represented by an encoder trained in a self-supervised fashion using pairs of adjacent segments assumed to be of the same speaker. The trained encoder model is then used to self-generate pseudo-labels to subsequently train a similarity score between different segments of the same call using probabilistic linear discriminant analysis (PLDA) and further to learn a clustering stopping threshold. We compared our model to state-of-the-art unsupervised as well as supervised baselines on the CallHome benchmarks. According to empirical results, our approach outperforms unsupervised methods when only two speakers are present in the call, and is only slightly worse than recent supervised models., Comment: Submitted to Interspeech 2022

We analyze energy spreading for a system that features mixed chaotic phase-space, whose control parameters (or slow degrees of freedom) vary quasi-statically. For demonstration purpose we consider the restricted 3~body problem, where the distance between the two central stars is modulated due to their Kepler motion. If the system featured hard-chaos, one would expect diffusive spreading with coefficient that can be estimated using linear-response (Kubo) theory. But for mixed phase space the chaotic sea is multi-layered. Consequently, it becomes a challenge to find a robust procedure that translates the sticky dynamics into a stochastic model. We propose a Poincar\'e-sequencing method that reduces the multi-dimensional motion into a one-dimensional random-walk in impact-space. We test the implied relation between stickiness and the rate of spreading., Comment: 16 pages, 13 figures

A new method for the classification of respiratory diseases is presented. The method is based on a novel class of features, extracted from pulmonary sounds, by parameterizing their spectrograms that are represented as surfaces, and by utilizing geometrical distortions defined with reference to these surfaces. This method yields a set of highly descriptive features for the analysis of pulmonary sound recordings. Furthermore, by combining these features with Mel-frequency cepstral coefficients, we introduce a powerful model for the automatic diagnosis of common respiratory pathologies. Compared with baseline methods, our model achieves superior results for binary and multi-class classifications of common respiratory diseases. Our new approach to the classification of one-dimensional signals is applicable to other signals in the context of their representations in combined spaces or manifolds.

Targets search and detection encompasses a variety of decision problems such as coverage, surveillance, search, observing and pursuit-evasion along with others. In this paper we develop a multi-agent deep reinforcement learning (MADRL) method to coordinate a group of aerial vehicles (drones) for the purpose of locating a set of static targets in an unknown area. To that end, we have designed a realistic drone simulator that replicates the dynamics and perturbations of a real experiment, including statistical inferences taken from experimental data for its modeling. Our reinforcement learning method, which utilized this simulator for training, was able to find near-optimal policies for the drones. In contrast to other state-of-the-art MADRL methods, our method is fully decentralized during both learning and execution, can handle high-dimensional and continuous observation spaces, and does not require tuning of additional hyperparameters.