Search

Large-scale sequence modeling has sparked rapid advances that now extend into biology and genomics. However, modeling genomic sequences introduces challenges such as the need to model long-range token interactions, the effects of upstream and downstream regions of the genome, and the reverse complementarity (RC) of DNA. Here, we propose an architecture motivated by these challenges that builds off the long-range Mamba block, and extends it to a BiMamba component that supports bi-directionality, and to a MambaDNA block that additionally supports RC equivariance. We use MambaDNA as the basis of Caduceus, the first family of RC equivariant bi-directional long-range DNA language models, and we introduce pre-training and fine-tuning strategies that yield Caduceus DNA foundation models. Caduceus outperforms previous long-range models on downstream benchmarks; on a challenging long-range variant effect prediction task, Caduceus exceeds the performance of 10x larger models that do not leverage bi-directionality or equivariance., Comment: ICML 2024; Code to reproduce our experiments is available at https://github.com/kuleshov-group/caduceus

Software-defined networking is considered a promising new paradigm, enabling more reliable and formally verifiable communication networks. However, this paper shows that the separation of the control plane from the data plane, which lies at the heart of Software-Defined Networks (SDNs), can be exploited for covert channels based on SDN Teleportation, even when the data planes are physically disconnected. This paper describes the theoretical model and design of our covert timing channel based on SDN Teleportation. We implement our covert channel using a popular SDN switch, Open vSwitch, and a popular SDN controller, ONOS. Our evaluation of the prototype shows that even under load at the controller, throughput rates of 20 bits per second are possible, with a communication accuracy of approximately 90\%. We also discuss techniques to increase the throughput further.

Future network services present a significant challenge for network providers due to high number and high variety of co-existing requirements. Despite many advancements in network architectures and management schemes, congested network links continue to constrain the Quality of Service (QoS) for critical applications like tele-surgery and autonomous driving. A prominent, complimentary approach consists of congestion control (CC) protocols which regulate bandwidth at the endpoints before network congestion occurs. However, existing CC protocols, including recent ones, are primarily designed to handle small numbers of requirement classes, highlighting the need for a more granular and flexible congestion control solution. In this paper we introduce Hercules, a novel CC protocol designed to handle heterogeneous requirements. Hercules is based on an online learning approach and has the capability to support any combination of requirements within an unbounded and continuous requirements space. We have implemented Hercules as a QUIC module and demonstrate, through extensive analysis and real-world experiments, that Hercules can achieve up to 3.5-fold improvement in QoS compared to state-of-the-art CC protocols.

Learning dynamics from dissipative chaotic systems is notoriously difficult due to their inherent instability, as formalized by their positive Lyapunov exponents, which exponentially amplify errors in the learned dynamics. However, many of these systems exhibit ergodicity and an attractor: a compact and highly complex manifold, to which trajectories converge in finite-time, that supports an invariant measure, i.e., a probability distribution that is invariant under the action of the dynamics, which dictates the long-term statistical behavior of the system. In this work, we leverage this structure to propose a new framework that targets learning the invariant measure as well as the dynamics, in contrast with typical methods that only target the misfit between trajectories, which often leads to divergence as the trajectories' length increases. We use our framework to propose a tractable and sample efficient objective that can be used with any existing learning objectives. Our Dynamics Stable Learning by Invariant Measure (DySLIM) objective enables model training that achieves better point-wise tracking and long-term statistical accuracy relative to other learning objectives. By targeting the distribution with a scalable regularization term, we hope that this approach can be extended to more complex systems exhibiting slowly-variant distributions, such as weather and climate models., Comment: ICML 2024; Code to reproduce our experiments is available at https://github.com/google-research/swirl-dynamics/tree/main/swirl_dynamics/projects/ergodic

Dynamic 2-[18F] fluoro-2-deoxy-D-glucose positron emission tomography (dFDG-PET) for human brain imaging has considerable clinical potential, yet its utilization remains limited. A key challenge in the quantitative analysis of dFDG-PET is characterizing a patient-specific blood input function, traditionally reliant on invasive arterial blood sampling. This research introduces a novel approach employing non-invasive deep learning model-based computations from the internal carotid arteries (ICA) with partial volume (PV) corrections, thereby eliminating the need for invasive arterial sampling. We present an end-to-end pipeline incorporating a 3D U-Net based ICA-net for ICA segmentation, alongside a Recurrent Neural Network (RNN) based MCIF-net for the derivation of a model-corrected blood input function (MCIF) with PV corrections. The developed 3D U-Net and RNN was trained and validated using a 5-fold cross-validation approach on 50 human brain FDG PET datasets. The ICA-net achieved an average Dice score of 82.18% and an Intersection over Union of 68.54% across all tested scans. Furthermore, the MCIF-net exhibited a minimal root mean squared error of 0.0052. The application of this pipeline to ground truth data for dFDG-PET brain scans resulted in the precise localization of seizure onset regions, which contributed to a successful clinical outcome, with the patient achieving a seizure-free state after treatment. These results underscore the efficacy of the ICA-net and MCIF-net deep learning pipeline in learning the ICA structure's distribution and automating MCIF computation with PV corrections. This advancement marks a significant leap in non-invasive neuroimaging.

The performance of automatic summarization models has improved dramatically in recent years. Yet, there is still a gap in meeting specific information needs of users in real-world scenarios, particularly when a targeted summary is sought, such as in the useful aspect-based summarization setting targeted in this paper. Previous datasets and studies for this setting have predominantly concentrated on a limited set of pre-defined aspects, focused solely on single document inputs, or relied on synthetic data. To advance research on more realistic scenarios, we introduce OpenAsp, a benchmark for multi-document \textit{open} aspect-based summarization. This benchmark is created using a novel and cost-effective annotation protocol, by which an open aspect dataset is derived from existing generic multi-document summarization datasets. We analyze the properties of OpenAsp showcasing its high-quality content. Further, we show that the realistic open-aspect setting realized in OpenAsp poses a challenge for current state-of-the-art summarization models, as well as for large language models., Comment: EMNLP 2023